SYLLABUS PAPER RESEARCH METHODS Unit-1 Meaning, types and process of research: Meaning - purpose types of research - Pure,

applied, historical, analytical, descriptive and experimental -Significance of research in social sciences - Process of research - Meaning -Scientific method - Induction and deduction. Unit - 2 Planning Research : Research Problem - identification, selection and formulation of research problem - Review of literature in the field of corporate management - Hypothesis - Meaning - Sources of hypothesis - Types of Hypothesis - Formulation and testing - Research design - Factors affecting research design - Evaluation of research design. Unit - 3 Sampling design: Census method and sampling method for investigation - advantages and disadvantages of sampling - principle of sampling - Essentials of a good sampling - methods of sampling probability and non-probability sampling methods - Selection of a sample - factors affecting the size of the sample - Biased sample sampling and non-sampling errors. Unit-4 Sources of Collection of Data : Sources of data - primary and secondary data - Modes of data collection - Analytical method - case study observation - survey method - Interview - Its purpose and importance - Types of interview - Preparation for an interview Effective interview techniques - Limitations of Interview - Schedule Its meaning and kinds - Essentials of a good schedule -Procedure for the formulation of a schedule - Questionnaire Meaning and types -

Format of a good questionnaire - Factors affecting the response to a questionnaire Advantages and limitations of schedules and questionnaires -pre-testing and its importance. Unit-5 Processing and analysis of data : Meaning - Importance Process of data analysis - Editing - Coding - Tabulation - Diagrams The process of interpretation - Guidelines for making valid interpretation - Scaling techniques Meaning - Importance - Methods of their construction. Unit-6 Report Writing : Role and types of reports - Contents of research report -Steps involved in drafting reports - Principles of good report writing -Referencing - Criteria for evaluating research reports / research findings. REFERENCES 1. John W. Best, Research in Education 2. Anderson et-al, Thesis and Assignment Writing 3. Goode and Hatt, Methods of Social Research 4. Wilkinson and Bhandarkar, Methods and Techniques of Social Research. 5. ICSSR, Training in Research Methodology in Social Sciences in India.

UNIT-I RESEARCH: CONCEPT, TYPES & PROCESS

In this unit, meaning of research, definitions of research purposes of research, types of research, significance of research, process of research, scientific method and its components are dealt. INTRODUCTION This is the world of knowledge. As knowledge is power, the desire for more has led to search for more knowledge. The need for knowledge, more knowledge, cutting-edge knowledge, continually improved knowledge and the like drives the search for the same ever green and even more expanding. Research and Development (R&D) activities of businesses have been responsible for development of new products, new processes and new applications. Thus research is an endeavour that contributes to development. Below, the meaning, purpose, types, significance and process of research, particularly, in business sciences, are dealt with. 1.1 MEANING AND DEFINITIONS OF RESEARCH Research = Re + Search. That is search after search and search. Why? The purpose of such repeated search is to thoroughly understand the problem or issue or phenomenon and find the apt and effective solution to the problem or strategy to deal with the issue or the phenomenon. Such solution or strategy adds to our stock of knowledge in dealing with the problem or the issue. In common parlance, research refers to a search for knowledge. Research simply put, is an endeavour to discover answers to problems (intellectual and practical) through the application of scientific method. The Websters International Dictionary gives a very inclusive definition of research as as careful, critical inquiry or examination in

seeking facts or principles, diligent investigation in order to ascertain something". The 20th Century Chamber Dictionary defines research as: a careful search or systematic investigation towards increasing the sum of knowledge. Research is essentially a systematic enquiry seeking facts through objective, verifiable methods in order to discover the relationship among them and to deduce from them broad principles or laws. Research is really a method of critical thinking. It comprises defining and redefining problems, formulating and hypothesis evaluating or suggested making solutions, collecting, organizing data,

deductions and making conclusions and at last, carefully testing the conclusions to determine whether they fit the formulated hypothesis. Thus the term 'Research* refers to a critical, careful and exhaustive investigation or inquiry or experimentation or examination having as its aim the revision of accepted conclusions, in the light of newly discovered facts. Research is the power of suspending judgment, with patience, of mediating with pleasures of assessing with caution, of correcting with readiness and of arranging thought with scrupulous pain, according to Francis Bacon. Francis G, Cornell defines research as, "the activity of collecting information in an orderly and systematic fashion". D. Slesinger and D. Stephenspn in the Encyclopedia of Social Sciences define research as "the manipulation of things, concepts or symbols for the purpose of generalising to extend, correct or verify knowledge, whether that knowledge aids in construction of theory or in the practice of an art". John W. Best defines Research, "as the systematic and objective analysis and recording of controlled observations that may lead to the

developments of generalisations, principles, or theories,, resulting in prediction and possibly ultimate control of events". Definitions of this sort are rather abstract. So a summary of some of the characteristics of research may help to clarify the spirit, meaning, and methodology of research. The summary of characteristic features of research follows: Research is directed toward the solution of a problem. Research is highly purposive, Research involves the quest for answers to unsolved problems. Research is characterised by patient and unhurried activity. Research is logical and objective. Research is carefully recorded and reported. Research emphasises the development of generalisation of principles or theories that will help in predicting future occurrences. Research is based upon observable experience or empirical evidence. Research generally involves a hypothesis or set of hypotheses concerning explanation of a phenomenon. Research demands accurate observation and description. Research involves gathering new data from primary or first hand sources, or using existing data for a new purpose. Research activities are more often characterised by carefully designed procedures, always applying rigorous analysis.

 Research requires expertise i.e. skill necessary to carry out investigation, search the related literature and to understand and analyse the data gathered. Research applies every possible test to validate the procedures employed, the data collected and the conclusions reached. Researchers are not only scholarly and imaginative, but also have highest integrity and patience.

Research is thus the process of a systematic and in-depth study or search of any particular topic, subject or area of investigation, involving collection, compilation, presentation and interpretation of relevant details or data. It is a careful search or inquiry into any subject matter, which is an endeavour to discover to find out valuable facts which would be useful for solving problems or improving existing solutions to problems. The research that involves scientific analysis would result in the formulation of new theories, the discovery of new techniques, modification of old concepts or a knocking-off an existing theory, concept or technique. Madhavi Shah writes; Research is a word which frightens a lot of people. It need not. As a matter of fact, it is essentially a state of mind - a friendly, welcoming attitude towards change. Research for practical people as teachers, supervisors, etc should arise out of the desire to do things better.

1. 2 PURPOSE OF RESEARCH Research has proved to be a powerful tool in driving the mankind towards development. That is why, the terms Research and Development (R and D) are now used as twins as Research &

Development (R&D). John W.Best observes that, "the secret of our development has been research pushing back the areas of ignorance by discovering new truths, which, in turn, lead to better ways of doing things and better products". Today's emphasis in almost all fields of life is 'creativity and innovation'. Any creative and innovative mind, is actually a research mind. As John W.Best rightly observed, we owe all our development, social, cultural, economic, business, education, etc to research in these areas. There is no development without research. Research, according to J.Lahiri, economises effort, prevents wastage, increases efficiency, and reacts to vitalize and dignify the work of the concerned". i) Progress and good life: The purpose of all research in progress and good life. Progress results if the space of ignorance is occupied by knowledge and wisdom. The latter are the results of good research. Knowledge and wisdom drive the mankind to live an orderly good life. ii) Development of scientific attitude: One of the purposes of research is to develop scientific attitude. Scientific attitude is one that asks 'Why' and 'How' and answers are found. This 'Know-why' and 'know-how' attitude nurtures talents and such intellectual talents are the great assets of society. iii) Creativity and innovativeness: One of the purposes of research is encouragement to creativity and innovation. New products, new processes and new uses are the means through which the world goes dynamic. A dynamic world is not possible without newness introduced every now and then in every walk of life. And this is possible only through creativity and innovation. Research kindles the creativity and innovative instincts of people and thus experiments on the possibility of

new things instead of waiting for the accidental and slow experience path to creativity and innovation. iv) Testing hypothesis and establishing theories: A very

important/ purpose of research is testing of hypothesis and establishing theories. As was already pointed out knowledge is power. That knowledge comes from testing hypotheses and establishing theories. v) Prediction and Control: Applied research has a great say in prediction and control-in almost all walks of human endeavour. Prediction is jumping into the future and the theories constitute the launch pad. Control looks for deviation between actual happening and predicted happening. In the process, the theories get reevaluated and redefined. vi) Purposive development: Development = Growth + Change, Growth is uni-scaled while change is multi-scaled. In the natural process development does take place through trial and error through casual observations, through actual exposure and the like. But this is evolutionary and time consuming. Revolutionary development takes fourth through discontinuous dichotomous change. charge or Research even is the seed of such which disruptive change new theories. Proven hypotheses become

contributes to purposive development. vii) Problem Solving: The purpose of any research is problem solving. What is a problem? Problem is deprivation or depreciation of something. Knowledge deprivation, efficiency deprivation, productivity depreciation, etc exist. How can these be solved? Research into the forces that cause deprivation and measures to contain them from causing

deprivation is needed. Thus, problem solving is a great purpose of research. viii) Schematic Evaluation: Research is also carried, out to systematically evaluate a process or practice or an organisation to know its strengths and weaknesses so that areas for improvement process can be identified. ix) Impact Analysis: Research is undertaken to assess the impact of certain measures or change introduced on relevant variables. Impact studies are useful for biological, social, business, economic and other areas of decision making. x) Methodological Improvement: Another purpose of research is improving research methodology itself. Developments in the field of measurement and scaling are immense. Whether these can be appropriately used in the case of particular research areas? To answer the question research needs to be done. Validation, revalidation and devalidation of methodological aspects thus constitute good piece of research. And this is one of the purposes of research. In fact, any research has a responsibility enrichment 1.3 TYPES OF RESEARCH There are different types of research. Table 1.1 gives a list of the types of research. A sort of bipolar classification of research types is attempted here. towards contribution to methodological

Table 1.1 Types of Research 1 2 3 4 5 6 7 8 Pure Historical Exploratory Descriptive Analytical Case study Longitudinal Experimental Applied Futuristic Explanatory Prescriptive Synthetical Survey Cross-sectional Desk

Each of these types of research is described below. 1.3.1 Pure and Applied Researches Pure research or fundamental or basic research is more theory oriented. It is a search for broad principles and postulates that build a new theory or revalidate a theory or revalidate a theory. Take Einstein's Relativity theory: E = MC2. This is a fundamental truth which he has established. Nuclear fission and fusion are fundamental researches. Maslow's need hierarchy theory is fundamental in nature. Pure research is highly abstract in its orientation. The findings are of fundamental importance that more applications are built over the same. Hence, the research gets the name basic or fundamental research.

Whether capitalistic economic theories applicable in socialistic set up? How liberalisation, privatisation and globalisation (LPG) can be modeled for the Indian economy? Whether the cultural and sociological foundations of the West and East conform to or contradict with each other? Is Malthusian theory of population correct? Does Keyne's General Theory of Income and Employment hold good? Researches on these lines are fundamental in nature. They look into certain very fundamental issues. Establishment of a new theory or evaluation of an existing theory is the thrust of pure research. A word of caution is needed at this stage. It is very difficult to brand a research as pure or otherwise. Because, every research does have an application side as well. Applied Research is problem solving oriented. Hence, its other name action research. While basic research deals with theory, action research is concerned with practice. Most research works fall in this area, because there are more problems around requiring studies on them for evolution of possible solutions. Theories when put to use, really don't work, because, theories have the comforts of assumption or the ivory tower sophistication of tools. In the actual field, these theories fail to work and that the need for research to operationalise the theories arises. Thus the birth of applied research takes place. The impact of interest rate lowering on the scale and pattern of investment, of the LPG Policy on aggregate production, consumption and distribution, the impact of PDS on amelioration of rural poverty, the role of SHGs in improving the lot of rural households etc are all applied research works. We can give endless list of topics applied research.

Mathuri Shah writes, research should help to narrow down the proverbial gap between theory and practice. The thrust of applied research is activating theories into practice. 1.3.2 Historical and Futuristic Researches Historical research peeps backwards into history or the past. Study of the past to discern patterns over there, impact of the same on the present, the evolutionary process and so o& in a sense most researches are historical, because they use data and information pertaining to the past. Actually data and information mostly pertain to the past. Hence researches depend heavily on the past. Recent past? Distant Past? Deep distant past? These are the relevant question right now. Perhaps historical researches concentrate on deep distant past. Enquiry into the trade, commerce, business and economy of ancient India (5000 - 3500 BC), the administrative system during the reign of Asoka etc are truly historical research pieces. Historical research requires extraordinary skills on the part of the researcher to live the distant past, to visualise the then environment, to analyse and synthesise the undercurrents and overtones of the past. Everything must be brought before the mind-age is an action-replay form. Futuristic research looks forwards into the future. How a particular problem is going to unfold, what will be its implications for the rest, what policy orientations are called for, etc are the thrusts of futuristic regime. How policies will shape up in the future? How human needs will shape up in the future? How organisations will shape up in the future? These are certain examples of futuristic studies.

1.3.3 Exploratory and Explanatory Researches Exploratory research explores into a virgin or less researched area. Formulation of problems, concepts and hypotheses is the thrust here. The testing or examining the problems/concepts/hypotheses is not attempted. Exploring issues for research is attempted. It is necessary that issues for research are well established so that a good start for dealing with the issues is possible. The purpose of exploratory research is formulation of problem or hypotheses and not testing of the same. Explanatory research tries to offer explanations or solutions to the problem. While exploring research helps formulate problems, explanatory research helps developing solutions. The thrust of explanatory research is hypothesis testing. 1.3.4 Descriptive and Prescriptive Researches Descriptive researches give a neat account of: What is observed? What relationship is found? What forces contribute to the problem? What factors offer solutions? The emphasis is on stating "as is, where is". It is widely used. A detailed account of the past, present and future, with elaborate data and statistical backing is given. Trend in global investment and trade, trend in GDP and BOP, the periodicity of the occurrence of a phenomena, etc could be the topics of study. Descriptive researches are passive as they simply present the situation. Perhaps simple interpretations and generalisations are appended. Description of pricing strategies adopted, product-launches effected, required salesmen qualities, criteria of successful advertisement copies, styles of leadership adopted by business executives, labour problems such as absenteeism and labour

turnover, sources of recruitment adopted, etc are thrusts of certain descriptive studies in the field of business management. Prescriptive Researches put thrust on classifying things into good and bad. It is normative. What are the better sources of recruitment? What are the right style of leadership? What pricing strategies are preferable? Like the above, the thrust of prescriptive researches goes. Prescriptive researches are problem-solving oriented. Besides, fathoming out the dimensions of a problem and factors complicating it, solutions are offered. The solutions might have been formulated and tested under simulated environments or simply intuitive based. Repositioning a product, dealing with employee-discontent, etc prescribe ameliorative solutions to tackle sickening problems. 1.3.5 Analytical and Synthetical Researches Analytical researches look for specific trees in a vast 'wood'. To analyse means to dissect to find the inner combinations and complexities, to see the layers and spirals, to discern the dominant and dull and the like. Most researches are analytical in nature. Analysing the factors, their co-relationship, their effectiveness or severity, their tendency to form into sub-groups, their reaction to external inputs, their ability to reach a new equilibrium when disturbed, etc are the thrusts of analytical research. The above pattern of analysis is common in almost all analytical research, be it in physical, biological, social and other sciences. Diversity in unity is the thrust, Synthetical researches concentrate on unifying things and events. How is a theory developed in physical science could be applied

in social science? How is a model developed in capitalistic order be applied to a command economic order? These and similar articulations are the hallmark of synthetical researches. Unity in diversity is the thrust. 1.3.6 Case Study and Survey Research Case study is a comprehensive study of an entity or unit, be it a person, u group, an institution, a geographical area, a community, a plant variety and the like. The emphasis is to study the evolutionary history, the dynamic present am: the evolving future of the study unit or the case. Case study is an intensive stuck of the case in all dimensions and in all its times. Total mapping is attempted. The features of case study, sources of data and phases of case study are given below: Features: i) Study of the whole unit, as the unit is defined; ii) study is intensive, deep and thorough; iii) all factors, general or environmental, factors and specific factors having a bearing on the case are extensively covered; iv) historical perspective, evolution over time and space, and present scenario are given full account; v) case must be something distinct either as the best, or the worst or the via media. It was reported that a devoted individual recently committed himself to unweed the river Cauvery for well over 7 months without a single day break, all alone, despite ridicules from the rest, to a stretch of 8 kms. This is a daunting task. The dedicatedness of the individual has won the attention of the authorities and everyone. He is a good case, a distinct entity, to be studied. (He had been rewarded by the Chief Minister). Sources of Data: For case studies, personal or organisational documents, life history or organisational history, speeches and news items, etc are the sources. Interviewing the person or persons gives a lot more data. Continuity of data and

depth and breadth of data are very important. Phases of Case Study: Choice of case for study, preparations to approach the unit of study, collection and recording of data and interpretation of data are the phases of case research. The utilities of case study are as follows: helps formulation of valid hypothesis as certain deep insight into the behaviour of the case unit is got; helps in taking samples and helps in framing and refraining the questionnaire or interview schedule Survey researches as opposed to case studies, study a plural number of entities. The purpose is to find the population features, like the mean value, deviation from that value etc. Survey may be census survey or sample survey. 1.3.7 Longitudinal and Cross-Sectional Researches Longitudinal researches involve studying behaviour of the event or entity over a period of time. Time series backdrop is much involved. Cross-sectional researches try to see at a fixed point of time, the behaviour of a cross-section of elements. 1.3.8 Experimental and Desk Researches Experimental research involves conducting test in a simulated or real space and time. The emphasis of most experimental researches is establishing cause and effect relationship, acquiring sensitivity of the dependent factor on an independent factor, keeping other independent factors controlled and finding the conditions under which reactions take place.

The main features of experimental researches are: i) isolation of factors into dependent, independent and catalyst, ii) replication of the experiment to ensure the reliability of the results and iii) measurement of the result, i.e., inputs, conditioning environment and output. The hallmark of experimental research is preciseness and exactness. Everything is structured, controlled, monitored, measured and reported. There are different types of experimental research. These are: i) natural or uncontrolled one as in the case of weather, astronomy or other natural phenomena where only observation of things or phenomena is done and results analysed and conclusions drawn; ii) the laboratory-situation research where a simulated environment is created with the input and conditioning variables manipulated to find the output behaviour; and iii) the field experiment where research is conducted in a social setting, the researcher having only a less manipulative power as far as the input variables and the conditioning factors. The first type of research is used in the case of totally uncontrollable phenomena like gravitational force, weather, astronomical or celestial events. The second type is used in physical, biological and psychological sciences. The third one is based in management, business and social sciences. There are several experimental designs randomised" design, like: completely

completely randomised block design, -Latin

square design, cross-sectional design, longitudinal design, etc. The Desk Research simply involves working with published data. No experiment is conducted. Data found in published sources back the research. .Analyses are make and conclusions drawn.

1.4 SIGNIFICANCE OF RESEARCH IN SOCIAL SCIENCES Social sciences refer to business, commerce, demography, psychology, sociology, etc. Social sciences directly involve people, as such. Research in social sciences arena deals with the behaviour of people in their different roles, such consumers, competitors, followers, producers, executives, salesperson, leaders, workers,

teachers, students, opinion-makers, etc. Research in social sciences deals with the systematic method of discovering new facts or of verifying old facts, their sequence, inter-relationship, casual explanations and the natural laws which cover them. The importance that social science research wields today is immeasurable and enlarging. As social, business and economic problems abound, the significance of social research gets enhanced as it provides workable solutions. We know the objectives of social research are elaborate. From these emanate the significance of research. The following points bring out the significance of research in social sciences. Problems solving is the thrust of most researches. Social problems are felt directly by people and that research by offering solutions to such problems ameliorates the conditions of people at large. Hence the significance of social research. Social research thrusts on societal behaviour, which is studied, analysed and steps needed to modify the same to achieve certain broad goals. All our social problems could be attributed to certain societal behaviour. So, by modifying the same in the right lines, social good is achieved. Development of methodology to deal with social issues is one of the contributions of social research. Executive stress, worker ethics,

leadership style, child labour, women illiteracy, drug addiction, labour absenteeism, etc are social issues related to organisations, comrn unities, and, such other social groups. To deal with these issues appropriate methodology is needed. Social research provides the same. Social research contributes to societal development. The

research develops scientific temper. Creativity and innovation are developed Basic and applied new knowledge is developed. All this adds to upgradation of society. Knowledge is power. And that power is powered by research. Formulation of new theories and reevaluation of already accepted theories are attempted by social research. There are several theories on leadership, motivation, human attitude and behaviour and so on. All these theories help designing suitable packages for societal behavioural upliftment. Social research is a tool for social planning, prediction and control. Any constructive action need to be planned, outcome predicted and deviation of actual from the desirable predicted outcome need to be controlled. Social research aids in designing appropriate models of social planning, prediction and control. Social research contributes to social welfare. Social research is generally normative emphasizing what is good for the society. By stating, what is and what is not good for the economy, for the industry, for the consumers, for the students, for the stock-market and the like, social research helps to contribute to social welfare. Social research catches the dynamics of social institutions and phenomena. Social institutions and phenomena are never static. These keep changing. To gauge the change research is needed and

such research helps in dynamically responding to social institutions and phenomena. 1.5 PROCESS OF RESEARCH The process of research refers to the series of steps involved in research, The following are the steps involved in research. i) ii) iii) iv) v) vi) vii) viii) ix) x) xi) xii) xiii) xiv) Research reconnaissance Problem identification Problem formulation Problem delimitation Review of literature Formulation of research questions Formulation of research objectives Formulation of Hypotheses Determination of data requirement Identifying source of data Determination of mode of conduct with the source of data Designing the instrument of data collection Deciding the instrument of data collection Design of sampling method and preparing the sampling frame xv) xvi) xvii) Choosing the sample units Pre-testing the instrument of data collection Pilot study

xviii) Finalization of instrument and method of execution xix) Actual survey or study and data collection

xx) xxi)

Data editing, transcription, coding and tabulation. Data analysis: i) Uni-variate bivariate and multivariate, ii) Prescriptive, descriptive and inferential.

xxii) Hypothesis testing and inference. xxiii) Report preparation: Organisation of ideas and plan of report, xxiv) Drafting, re-drafting and finishing touches to the report xxv) Final report preparation and submission of report

Since all the steps are dealt at length at appropriate places, discussion of the same is not attempted here. 1.6 SCIENTIFIC METHOD The term science is defined as an 'accumulation of

systematic knowledge'. The sole purpose of science is to understand the world. Basic to science is an intricate relation between theory and fact. Theory and fact arc intertwined. Theory is not speculative. Theory helps in predicting facts. The Chambers 20th Century Dictionary defines science as "knowledge ascertained by observation and experiments, critically tested, systematised and brought under general principles". Scientific method, must therefore, include any systematic, testable and replicable process or experiment by the end of which knowledge is ascertained, tested and systematised. Scientific method is objective, .formal, rigorous, measurable, repeatable, verifiable, validatable and reliable. 1.6.1 Forms of Scientific Method There are different forms of scientific method. These are: historical & futuristic, descriptive & prescriptive, pure & applied, deductive & inductive and so on. Yes, each type of research dealt

already in this chapter is premised on some scientific method, process or base. Those which are not already dealt are elaborated a bit here. Structural Method involves studying the entity or sub-entities or super-entities, the relationship among the entities, the forms and norms of the entities and the like. The 'what', 'which' and 'where' questions are addressed. The functional method involves studying the functioning, the roles, causative and reactive domains, etc of the entity studied. The 'why' and 'how' questions are addressed here. Deductive Method involves drawing conclusions on individual entities based on the general pattern found. That is going from general to particular is deductive method. Democratic structure leads to better participation. A particular leader is a known democratic. So, the conclusion is he gets a higher level of participation from his colleagues or subordinates. This is deductive method. Inductive Method goes to generalise, from particularised or sample behaviour. This is the converse of deductive method. From individual.you move to the universe. 1.6.2 Elements of Scientific Method Concepts and hypothesis are the basic elements of scientific method. Concepts are symbolised phenomena that are studied. Concepts are themselves are not phenomena, these are symbols used to refer the phenomena. The concepts are abstraction and represent certain aspects of the empirical world that is studied. A concept is abstracted from sense impressions or percepts. Concepts must help clear

communication. They must not lead to any vague feeling. To ensure this precise use of words to mean relevant phenomena is needed. Concepts are developed from shared experience. Hypotheses are propositions which can be put to tests to determine their validity. Hypothesis state what we are looking for. If take a theory we may deduce number of relationships among the factors and forces connected to the theory which are not yet tested. At that moment, it is not possible to state whether are or not those relationships correct. In that form, the untested relationships constitute hypotheses. A testing scheme is needed to verify the hypotheses. Whatever the test unfolds into, a hypothesis is a question put in such a way that an answer of some kind can be forthcoming. 1.6.3 Induction and Deduction Induction is defined as, 'reasoning from particular cases to general conclusions' by the Chambers 20th Century Dictionary. Most researches are of this type. What is found true for a sample is generalised as true for the population as well. This method has a sound reasoning base and hence is, widely applied. Induction is concerned with first obtaining data, discovering the truth and establishing relationships regarding the particular cases concerned. If these are found correct and operating, these are generalised. Induction lays stress on establishment of the material truth of matters pertaining to the universe or population based on sample observations. In mathematics, the proof for the Binomial Theorem is based on induction method. A study of, say, a dozen, randomly chosen family run businesses reveals that the decision-making process is centralized and less transparent. And, it may be concluded that decision process is centralised and less transparent in the case of family-run businesses. Induction involves, moving-to the 'unobserved' many from the 'observed' few.

There are a few types of induction. Perfect induction and intuitive induction are these. Perfect induction involves establishing a universal proposition by an exhaustive enumeration of all the instances of the category covered by the proposition. But, by the yardstick of, 'moving to the 'unobserved' from the 'observed', perfect induction is not induction at all, as everything is observed and there is no need to move to the 'unobserved'. But a point is to be kept in mind. What is observed are individual entities and what is concluded is about the 'collective whole' or the 'totality*. Therefore, there is a definite more from 'observed' to 'unobserved'. The merits of induction method are: this is the only method of drawing conclusion in certain cases; it does not proceed with any assumption; mostly adopted in research and in practical situation too; sampling and induction go together. The dements of induction method are: lack of quality data; the difficulties in collecting data; the representativeness of the sample both as to time and space; fallacious conclusions such as: A sample of well-to-do students indulge in; so all well-to-do students indulge in. this may be true. 1.6.4 Deduction method All humans are mortal. A universal truth. Ram is a human. So, Ram is mortal. This, deduction involves, moving from general to particular. Members of upper-occupational strata experience less unhappiness and worry and are subject to more formal controls than members of lower strata. This is a principle. It is deducted from the principle that: there would be more happiness in the wealthy families than in the poor families.

Deduction method helps to formulate happiness as is referred to above by extending a scenario into an analogous scenario. To test the hypothesis so formed, induction method may be adopted. Hence, the use of both deduction and induction in research. Deduction is concerned with what is the material truth of the premises. It does not go beyond that premises. Conclusions based on deduction are more certain, than in the case of induction. QUESTIONS 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Define research and bring out its features. Explain the research and its purposes. What are the types of research? Explain What is a case study? How does it differ from survey research? Explain pure and applied research. Bring out the significance research in social sciences. Explain the process of research. What is scientific method? Explain its features. Explain deduction and induction. Bring out the need for research in business.

REFERENCES: 1. Elements of Educational Research - S.P.Sukhia & D.V. Mehrotra. 2. Methods in Social Research - Goode & Hatt. 3. Research Methodology - P. Saravanavel. 4. Research Methods in Commerce - D. Amerchand (Edr)

*******

UNIT - II RESEARCH PROBLEM, HYPOTHESIS AND RESEARCH DESIGN In this unit Planning of Research, Research Problem,

identification, selection and formulation, Review of literature in, the field of corporate management, Hypothesis Meaning, Sources of hypothesis, Types of Hypothesis, Formulation and testing of hypothesis, Research design. Factors affecting research design and Evaluation of research design are dealt. INTRODUCTION The first step in research is formulating the research problem. Review of literature gives both ideas as to research problems and also how earlier research works have been conducted. Both problem formulation and review of literature enable a researcher to formulate the objectives of the research study and from the objectives to formulate hypotheses for the study. Works upto this phase and after wards need to be given a definite shape and order in the form of a research design. Hence in this unit research problem, review of literature, hypothesis and research design are dealt. 2.1 RESEARCH PROBLEM A research problem is defined by F.L. Whitney as, "that unsolved difficulties". Stefan Nowak defined a research problem as "a certain question or set of questions to which the research is to provide an answer". In this section we shall deal with research problem in terms of i) problem identification, ii) conditions as to a research problem, iii) basic concepts in identification of research problem, iv) sources of research problem, v) types of research problem and vi) problem form.

2.1.1

Problem Identification Identifying a research problem is the first and foremost step in

any research process. A researcher has to devote considerable time and attention for this. A research problem refers to some deprivation or shortcoming or a gap in knowledge that a researcher experiences in a theoretical or practical situation and wants to find a solution for the same. Research really begins when the above referred situation demanding a solution is perceived by the researcher within the realm of general topic selected by hum. The general area of interest here refers to the range of the subject matter within which he has to see and pose a specific problem for research. So, the formulation of the topic into a research problem is. really, the first step in a scientific enquiry. The general area of study may be suggested either by some practical concern or some scientific or intellectual curiosity, which in turn may be a general subject of which little is known or a particular phenomenon which may be probed more deeply. In fields which have a highly developed theoretical base, the researcher may want to put specific predictions and expectations to test against the much established theory. The selection of a topic for research is only half a step forward. Generalization of a- topic is of no relevance to the researcher in as much as it does not help him perceive, secure and organize his data. Hence the need to formulate a very specific problem. A problem defines the goal of the researcher in clear terms. It is said that, "if you start from nowhere, you will generally reach there". Without a goal, i.e., without a specific problem posed before him, the researcher will be searching for a needle in a haystack. A well defined goal for a researcher is like a well-lit road before a traveler in the pitch darkness of night. The more pin-pointed the topic, the more it comes into focus

before the researcher. Thus, it is clear that the topic should be so formulated that its scope could be progressively narrowed and then specific challenging questions, emerge in simple terms. This is essentially the first step in a scientific enquiry. 2.1.2. Conditions as to Research Problem

Usually a research problem selection depends on the following conditions on the practical side. (i) There should be a pressing need for a factual evaluation of a programme, information pertinent to policy making or marketing planning or necessity for evolving a suitable solution to a dire problem the company faces. (ii) It is worth selecting a research problem which tends to touch upon or has a bearing on organisational values. Market conditions, financial viability of the researcher and the time at disposal play an important and subtle role in determining the researcher's preference in the selection of a topic, for inquiry. The choice of research topics is also affected by premia in respect of work done in the various fields. (iii) There must be an individual, group or an organisation to whom the problem can be attributed. They in turn belong to an environment which is defined by values of the uncontrolled variables. (iv) There must at least be two courses of action to be pursued. A course of action is defined by one or more values of the controlled variables.

(v)

There must be at least two possible outcomes of the course of action, of which one should be preferable than the other for it is that which the researcher is looking for his objective in doing research.

(vi)

The course of action must provide some chances for attaining the objective, with unequal efficiencies for the desired outcome.

(vii)

Further, the individual or organisation also can be said to have a problem if it does not know what course of action is best of the available alternatives.

The formulation of the problem also aims in making various components of .the problem explicit. A problem well framed is half solved. It would seem easy to pose a problem for research, but in reality it is not so. Even Charles Darwin in his Origin of Species has testified to this by saying "Looking back, I think it was more difficult to see what the problems were than to solve them". There is another point to remember here. Although every problem involves a question or a series of questions, not every question qualifies as a scientific problem. 2.1.3 Basic Concepts in Problem Identification Guidance in identifying a problem of research is afforded when we go through R.L.Ackoffs. The Design of Social Research wherein he visualized the five components of a research problem. In other words, research problems may be identified on the basis of these five concepts, which are in fact touchstones for a researcher. They are: (i) There must be an individual or a group which has some difficulty or a problem.

(ii)

There must be some objectives to be attained at the end. There can be no problem when there is no want,

(iii)

There can also be no problem, if there is no choice. So, alternative means or courses of action or instruments to meet or pursue the objectives must be available.

(iv)

The researcher must provide answer for the relative efficiency of the alternative means that are available to pursue the objective.

(v)

Another component of research is that there must be some environments to which the difficulty or problem pertains.

2.1.4

Sources of Research Problem Research problems may be selected from the following sources:

(i) (ii) (iii) (iv) (v)

Theory of one's own interest: Daily problems Technological developments Unexplored areas Discussions with expert authorities including the supervisor or research advisor. These sources are explained below: The research may select a problem for investigation from a given

theory in which he has considerable interest. In such situations the researcher must have thorough knowledge of that theory and should be sufficiently inquisitive to explore some unexplained aspects or assumptions of that theory.

Research problem can also be selected on the basis of daily experience of a researcher. Every day problems constantly present something new and worthy of investigation and it depends on the sharpness of the researcher's intellect to tend his daily experiences into a research problem. Technological developments in a fast changing society are constantly brining forth new problems and new opportunities for research. What is the impact of a changed information technology on the existing socio-economic set up? What is the impact of biotechnology in our economy? A good starting point either for selecting or for learning about-a topic is some one who is well acquainted with the topic, perhaps some one who has been involved in research in the topic. 2.1.5 Types of Research Problem Social science research problems may be of three distinct varieties differing in form, content, and mode of verifications. These are (i) empirical problems, (ii) analytical problems, (iii) normative problems. A brief on each is given below: (i) Empirical Problems : When social researchers answer

questions or problems on the basis of what they come to know through their sensory organs, these are expressed as empirical problems. Social researchers base their conclusions and findings on what they perceive or observe or sense in order to verify, approve or reject hypothesis. (ii) Analytical Problems : Analytical problems dissect into the factors causing a problem, the influence of these factors, the

factors that offer a solution, the efficacy of these factors and so on. (iii) Normative problems : Normative problems are questions whose answers depend primarily on value judgments. Value judgments are statements of what is desirable, preferred, moral, imperative or obligatory. These may take either an evaluative or prescriptive form. 2.1.6 Problem Formulation Problem formulation is structuring the problem such that it is focused, pointed and absolutely clear. Problem formulation has certain stages, definition of problem, and delimitation of the problem and justification of the problem. 2.1.6.1 Problem definition

Problem definition gives a concise and clear statement of the problem in terms objectives of the study, research questions of the study and the hypothesis of the study. It is important to define and elucidate the problem as a whole and further define all the technical and unusual terms employed in the statement. By this the research worker removes the chance of misinterpretation of any of these crucial terms. The definition helps to establish the frame of reference with which the researcher approaches the problem. 2.1.6.2 Delimitation of the problem

Delimitation involves earmarking the scope of the investigation. determines the boundaries of the problem. This delimitation mentions

the functional and geographical limits of the study i.e., whether the study will be covering a certain town, a district, a region, a state on a country. Again it is important to mention as to what subjects will constitute the sample of the study and how they will be distributed over the institutions, geographical areas or time intervals. 2.1.6.3 Justification of a Research Problem

Before the proposed research problem can be considered appropriate, several searching questions should be raised. Only, when those questions are answered in the affirmative can be problem be considered a good one. i) Is this type of problem can be effectively solved through the process of research? Can relevant data be gathered to test the theory or find the answer to the problem under consideration? ii) Is the problem significant? Is an important principle involved? Would the solution make any difference as far as theory or practice is concerned? iii) Is the problem a new one? Is no answer already available? Ignorance of prior studies may lead a researcher to spend time needlessly on a problem already investigated by some other researcher. iv) Is research on the problem feasible? After a research work has been evaluated, there remains the problem of suitability for a particular researcher.

2.2

REVIEW OF LITERATURE Once the research problem is identified and its delimitation

made, review of literature is carried out. Sometimes literature review helps in problem identification. But this may be referred to as research reconnaissance. Review of literature means a review or study of literature relevant to the study taken up by the researcher. The literature may be i) research reports, ii) research papers, iii) reference books, iv) periodicals covering issues relevant to the topics of research, v) government policy notes and reports on issues relevant to the topics, vi) relevant statutes, case laws and legal pronouncements, vii) scholarly text books, etc. The literature may be published or unpublished (but must be authenticative), local, national and global, individual or organizational, periodic or one-time and so on. 2.2.1 Objectives of review of literature Review of research report is done to know what research works have already been done on this and related topics or fields, the methodology adopted by them, the hypotheses formulated and tested by them, the findings and conclusions, the listed scope for further research, and so on. Similarly review of research papers is intended to know from the papers the thrust issues, the hypothesis the tool of data collection, the methodology, the findings and the implications of them for the research for his current work. Review of literature helps understanding the present research in a better way. In fact, review of literature beginning with a research for a suitable topic for research and continues until final stages of the current research work. 2.2.2 Purpose of Review of Literature

The following are the purposes of review of literature:

i)

To avoid duplication of research work, in case an almost similar work as the present one has already been done by some one.

ii)

To adopt an alternative method of study in case an almost similar work as already done has to be for some reason carried out.

iii)

To fill up gaps in research by incorporating objectives not thoroughly probed by earlier researchers.

iv)

To effectively, formulate hypotheses in the light of findings of earlier research works or policy pronouncement of governments or statutory development.

v)

To test the efficacy of suggested courses of action in earlier works.

vi)

To ward of methodological problems or troubles that might have been suffered by earlier researchers.

vii)

To refine the methodology of research in the light of suggested improvement by or for earlier works.

viii)

To add new dimensions of research in a study based on recommendations of earlier research works or scholarly research papers or legal pronouncements or investigations or other reports.

ix)

To have an idea of the diverse tools of analysis used by others and that could be used for the present research.

x)

To build a theory, review of literature is needed. Fitting the findings of the present work with the findings of earlier work, and on that basis evolving a theory becomes possible.

xi)

Jonatham Anderson writes that it is useful initially to survey a field of work from the general to the specific or to work from tertiary and secondary sources of information to primary sources.

xii)

Review of earlier research works also helps in developing on impressive structuring of research report. Based on knowledge gained on this aspect from the structures adopted in earlier works, right structuring of report is worked out.

2.2.3

Review of works on corporate management Review of research works on corporate management relating to

one's topic of research is needed. Research works on marketing management, operations financial management, management, production personal management, systems management,

management or information management need to be fully reviewed depending on one's topic of research. 2.3 HYPOTHESIS: MEANING A hypothesis is an assumption about (i) the relationship between/among variables or (ii) about the level of influence of independent variables on the dependent variable or (iii) about the value of population parameter. Thus relationship between income and brand preference, between education and habit of savings, between cost of capital and capital structure, between motivation and productivity, etc., are hypotheses of the first type. The influence of

sales on profit, of profit on value of business, of reward on performance and such other things are the second type. The third type involves putting values on population parameters such as the mean monthly salary of MBAs who graduated from a particular university in the first year of their employment is Rs. 18000 in 2003 or on an average MBA graduates wait for 6 months after graduation to get job or 40% MB As get fat jobs and so on. In the words of George A. Lundberg "A hypothesis is a tentative generalisation, the validity of which remains to be tested. In its most elementary stage the hypothesis may be very hunch or guess, which becomes the basis for action or investigation". Similarly, according to another author, "A hypothesis is a proposition, condition or principle which is assumed, perhaps without belief, in order to draw out its logical consequences and by this method to test its accord with facts which are known or may be determined". Goode and Hatt have defined a hypothesis, "a proposition which can be put to test to determine validity". Rummel and Balline say, "A hypothesis is a statement capable of being tested and thereby verified or rejected". William C. Emory says "when propositions are formulated or empirically tested they are called hypotheses". Propositions are combinations of concepts designated by statements that may be judged true or false if they refer to observable phenomena. Concepts are abstractions from reality and are designated by words or phrases. According to M.H.Gopal, "a hypothesis is a tentative solution posed on a cursory observation of known and available data and

adopted provisionally to explain certain events and to guide in the investigation of others. It is, in fact, a possible solution to the problem". 2.3.1 Types of Hypothesis Hypotheses are of different types. These are put here in dichotomous groups below: Abstract hypothesis Descriptive hypothesis Null hypothesis Explorative hypothesis Measurable hypothesis Relational hypothesis Alternative hypothesis Explanatory hypothesis

Abstract Vs Measurable Hypothesis : Hypotheses are classified as abstract and measurable. Abstract hypothesis is imprecise and testing of the same is difficult. Measurable hypothesis is precise and can be tested. Descriptive Vs Relational Hypothesis : Descriptive hypothesis assumes certain parametric values that describe population features. Relational hypothesis deals with relationship or association between variables. Null Vs Alternative Hypothesis : Null Hypothesis is a hypothesis which assumes that there is no difference between .two values such as population means or population proportions and in case there is any difference that is attributed to sampling fluctuations and is unimportant.

In many situations, we set up a null hypothesis simply for the purpose of rejecting it. For example, to know whether fertilizer manufactured by one firm will produce a higher yield of corn than fertilizer B manufactured by a second firm, the null hypothesis concerning the yield of corn of fertilisers is stated as: "There is no difference between the yield of Com from fertilizer - A and the yield of corn from fertilizer - B. Such hypotheses .are referred to as null hypotheses and are denoted by Ho. The researcher's intention is rejecting it with ample evidence. A hypothesis that differs from a null hypothesis is called an alternative hypothesis. As the name implies, an alternative hypothesis is a statement which will be accepted when the null hypothesis is rejected. An alternative hypothesis is denoted as H1. Regarding the yield of corn of two fertilizers A and B, the yield of corn from fertilizer A is different from yield of corn from fertilizer B, is one form of alternative hypothesis. The null hypothesis is that there is no difference between the mean yields of corn (H1H2) of two fertilisers A and B. In symbols, it may be written as: Ho Ho H1= H2 H1- H2 = 0 Or

The alternative hypothesis is stated that the mean yields of corn of two fertilizers are not equal. In symbols, it may be written as : i) ii) Or H1 H2 (Two tailed test) Either H1 H1 > H2 H1 H1 < H2 (one tailed test or upper tail test) (one tailed test or lower tail test)

Explorative Vs Explanatory Hypothesis : Explorative hypothesis tries to explore the possibility of relationship between certain variables

or influence of one over the other. Explanatory hypothesis explains the relationship if any between variables, the extent of influence of one factor over the other and the like. 2.3.2 Sources of Hypothesis There are diverse sources of hypothesis. First, an explorative research work might lead to the establishment of hypothesis. Second, the environment is a source of hypothesis, because environment portrays broad relationship across factors which form the basis for drawing an inference. Third, analogies are a source of hypothesis. The term analogies refer to parallelism. Though human system and animal system are different, there is some parallelism. That is why medicines are tried first on rats or monkeys then used for human consumption. So, hypothesis on animal behaviour can be done based on proven behaviour of human and vice versa. Similarly, between thermodynamics and group dynamics, biological system and social system, nervous system and central processing unit of a computer, parallelism can be thought of and spring hypotheses therefrom. Fourth, previous research studies are a great source of hypotheses. That is why review of literature is made. Fifth, assumptions of certain theories become a source of hypothesis. Similarly, exceptions to certain theory are ground for new hypotheses. Sixth, personal experiences and experiences of others are another source of hypotheses. Everyone encounters numerous experiences in day to day life in relation to one's avocation. From these glimpses of hypothetical relations between events, variables, etc emanate. These are, therefore, bases for establishment of possible hypotheses. Seventh, social, physical and other theories and laws provide for hypotheses. Newton's laws of motion might be a Source of hypotheses, in social science, say behaviour and reward and the like. Finally, for the research mind, the whole universe is a source of hypotheses. Yes. the searching mind fathoms out new hypotheses from seemingly events of insignificance.

2.3.3

Formulation and Testing of Hypothesis Hypothesis must be stated in the measurable and verifiable

farm and in the form null and alternative hypotheses so that the same can be tested. Several other steps are needed for testing hypotheses. 2.3.3.1 Formulation Suppose you want to test the equality of-mean production of three machine types. Then the null hypotheses is: H1 = H2 = H3 and alternative hypothesis is: Hi = Hj [That is the Ho says that productivity is same for all machines and H1 says at least two machines may be 1 and 2, or 2 and 3 or 3 and 1 differ in their productivity. For this ANOVA test is made. Suppose you want to test the equality of standard deviation of returns on two securities. The H0 1 = 2 and Hi can be either of the types: H1 : 1 : 2, 1 < 2 2 Here, Z test is used. 2.3.3.2 Level of Significance

As a matter of practical convenience, experiments and research workers have used probability levels for deciding whether to accept or to reject a null hypothesis. These probability levels are called levels of significance. It is defined as the probability of rejecting a null hypothesis, which is in fact true. In fact, two most commonly used levels of significance are 0.5 (5%) level and 0.01 (1%) level. These levels correspond to the confidence coefficients of 0.95 (95%) and 0.99 (99%) respectively. The level of significance is the maximum probability of rejecting a correct null hypothesis

Suppose a sample of 100 items is taken from a population with mean 136. The mean the sample works out 128 and the standard error of mean to 2.5. Here, we formulate the null hypothesis that there is no difference between sample (mean and population mean and whatever difference in these may be attributed to sampling fluctuations. Now we specify 0.05 probability at which we will reject the null hypothesis. This figure means that if the difference between the sample mean and population mean is such that it would occur only 5% or less than 5% of the times, the null hypothesis would wrongly be rejected. Here, the researcher runs the risk of making wrong decisions only in 5 out of 100 cases. In other words, the correct null hypothesis would be rejected in 5 out of 100 cases. If we specify 0.01 (1%) level of significance, this figure (0.01) implies that if the difference between the sample mean and the population mean is such that it would occur only 1% or less than 1% of times, the null hypothesis would he wrongly rejected. Here, the researcher runs the risk of making a wrong decision of rejecting the true null hypothesis only in one out of 100 cases. 2.3.3.3 Regions of Acceptance and Rejection

The level of significance is mainly employed in decision-making problems because it distinguishes between two regions known as region of acceptance and region of rejection. The region of acceptance may be defined as a range of values such that if the sample statistic falls in the range, the null hypothesis is accepted. On the contrary, the region of rejection may be defined as a range of values such that if the sample test statistic falls in this range, the null hypothesis is rejected.

2.3.3.4

Standard Error

The standard error is an indispensable tool in the kit of a research, worker; because it is used in testing the validity of statistical hypothesis. The standard deviation of the sampling distribution of a statistic is called the standard error of a statistic. The standard error is important in dealing with statistics (measures of samples) which are normally distributed. Here the use of the word "error" is justified in this connection by the fact that we usually regard the expected value to be true value and the divergence from it as error of estimation due to sampling fluctuations. The term standard error has a wider meaning than merely the standard deviation of simple sampling because of the following reasons; 1. The standard error is mainly employed for testing the validity of a given hypothesis. Mostly two levels of significance (0,05 and 0.01) are used for testing the validity of hypotheses. At 0.05 level of significance, if the difference between expected and observed values is more than 1.96 standard errors, the difference is considered significant Such a difference could not have arisen due to sampling fluctuations but due to the operation of other causes. On the contrary, if this difference is less than 1.96 standard errors, it is considered not significant, arising due to sampling fluctuations. In this case, the result of an experiment supports the null hypothesis. Similarly at 0.01 level of significance if the differences between observed and expected values is more than 2.58 standard errors, the difference is considered significant. Such a difference could have arisen not due to sampling fluctuations but due to other causes. On the other hand if the difference is less than 2.58 standard errors, it may be attributed to sampling fluctuations and can be

ignored. In this case, the result of the experiment supports the null hypothesis at 0.01 (1 %) level of significance. 2. The standard error may fairly be taken to measure the unreliability of the sample estimate. The greater the standard error the greater the difference between observed and expected values and greater the unreliability of the sample estimate. On the other hand, the smaller value of the standard error, the smaller the difference between observed and expected values and greater the reliability of the sample estimate. The reciprocal of the standard error may be regarded as a measure of reliability. The reliability of an observed proportion varies as the square root of the number of observations on which it is based. The larger -the sample size, the smaller the standard error and the smaller the sample size, the larger the standard error. In case, we want to double the precision the number of observations in the sample should be increased four times. 3. The most important use of the standard errors is in the construction of confidence intervals within which parameter values are expected to fall. For a sufficiently large sample size, the sampling distribution tends to be normal. Therefore at 5% level of significance, the population mean is expected to lie within the interval of: X + 1.96 standard error of mean. Similarly at 1% level of significance, the population mean is expected to fall within the interval of: X + 2.58 standard error of mean. In the same way, other confidence intervals may be constructed.

2.3.3.5

Two-tailed and One-tailed Tests

Depending on choice of alternative hypothesis, tests are classified into two-tailed and one-tailed tests. Below, we highlight these tests in detail: i) Two-Tailed Test : When two tails of sampling distribution of

the normal curve are used, the relevant test is called two-tailed test. Here, we are concerned with values that are higher as well as lower than the hypothetical value. For example, we may like to find whether the sample mean is significantly different from the hypothetical means or not. In the language of statistics it may be expressed as: Ho ; H = a specified value H1 ; H a specified value This is either, H > a specified value or H < a specified value

It may also be interpreted as that either very high values or very low values of sample mean (X) tend to support the alternative hypothesis (H1) and cause the rejection of null hypothesis (Ho). Therefore, we have two rejection regions in a two-tailed test. If it is decided to use 5% level of significance for a two-tail test, the null. Hypothesis would be rejected if the value of the test statistic is equal to or less than -1.96 and greater than or equal to +1.96. See the graph depicting two tailed test.

ii)

One-Tail

Test:

When

only

one

tail

of

the

appropriate

distribution is used, the lest is described as one-tail test. Here we are concerned only with values that are lower or only with values that are higher than the hypothetical value. For example, we may wish to determine whether the population mean is greater than a specified value i.e a hypothetical value. In statistical language, it may be represented as; Ho ; H = specified value H1 ; > a specified value It implies that very high values of sample means tend to support the alternative hypothesis (H1) and cause the rejection of null hypothesis (Ho). Here, we are not concerned with low values of sample means, therefore, we have one rejection number in one tail-test. Suppose, it is decided to use 5% level of significance for one tail-test, the null hypothesis (Ho) would be rejected if the value of test statistic is greater than + 1.645. See the graph depicting one tail-test.

2.3.3.6.

Type -1 Error And Type - II Error

In testing hypothesis, there is a possibility of making a wrong decisions because the decision is based on sample data. If the hypothesis, Ho is true, we accept and if the hypothesis is wrong, we reject it then we will be making a right decision. If the hypothesis, Ho is true and we reject it, we will be committing an error of type-1. On the other hand, if the hypothesis is false and we accept it, we will be committing an error of type-11. Type 1 error is committed when we reject a null hypothesis - Ho by marking differences significant when actually no such difference exists. In fact, null hypothesis is true but the sample data lead to the rejection of null hypothesis. Thus type-I error may be defined as the probability of rejecting a Ho, which in fact, is true. It is represented by the Greek letter

(called alpha), which is also the level of significance.

Similarly, type-II error is committed when we accept a Ho by making difference not significant when actually difference exists. In fact Ho is false but sample data lead to the acceptance of hypothesis. In other words, sample data do not seem to disagree with the hypothesis. Thus type-II error may be defined as the probability of

accepting a null hypothesis which is false. It is denoted by Greek letter (called beta). i) For a given sample size, the probability of an error of second type is determined once the probability of an error of first kind is decided. ii) If we want to decrease the probability of an error of first kind, the probability of type-II error is correspondingly increased. iii) If you want to reduce type-II error keeping type-I error at a fixed level, the size of sample must be increased. Thus, the four situations that arise in test procedure are described in the following table: Decisions State of Hypothesis H0 - True H0 Accepted H0 Rejected Correct (No Error) Type-II, Error Correct decision (No ) Error) 2.3.3.7 Procedures decision H0 - False Type-I, Error

The hypothesis test procedure is listed down here. First we have to state the null hypothesis. Second we have to state the appropriate alternative hypothesis. Third we have to state the level of significance. Fourth, we have to calculate standard error of relevant statistic. Fifth, we have to calculate the test statistic depending on the problem. Sixth, if need be "degrees of freedom" is to be chosen.

Seventh,

get

critical

values

from

appropriate

probability

distribution. Eighth, if the calculated test statistic (as in step 5 above) is less than critical value read from table, null hypothesis is accepted. Otherwise it is rejected. 2.3.3.8 Large Vs Small Samples vis-a-vis Distributions used

Statistically a sample is large when the size exceeds 30. In other words, when n > 30 it is larger same; when n < 30, it is small sample. When n > 30, i.e., for large samples normal distribution is generally used and for small samples t distribution is used in respect of testing of means, correlation coefficient, regression coefficients, etc. For testing of proportions Z test is used generally. For testing of mean and difference between means, normal distribution is used even in the case of small sample, if population standard deviation is known. In other words, t distribution is used only when population standard deviation is not known and the sample is small. 2.4 PARAMETRIC TESTS Parametric tests assume that a particular theoretical

distribution depicts a situation. It is concerned with estimating population parameters like means, proportions, regression coefficients etc. or testing their equality to an assumed value. 2.4.1 Testing for mean Here testing of equality of population mean to a hypothesised value of the same is done. Three test situations are involved.

i)

Large sample test of mean An example is given to explain this one. A foreign institutional investment house puts that high rated

debts return in India is 14%. A random sample of 100 AAA bonds gives a return of only 12.8%. The standard deviation of population return is known to be 4%. Test appropriate null hypothesis against suitable alternative hypothesis. Solution i) Two tailed test Null hypothesis: Ho: = 14% (Where = population mean return) Alternative hypothesis: H1 14 Alpha = Level of significance = 5% N = 100 and X= 12.8% Standard deviation = 4% Standard error of mean = SEX = /N = 4/100 = 0.4 Test statistic = Z - (X - )/SE = (12.8-14)/. 4 = -3 = 3 The normal distribution table value for alpha of 5% is 1.96 (for two tailed tests). The null hypothesis is to be rejected as the calculated Z = 3 is > critical value =1.96 [Note, even if population standard deviation is not known here, you can take sample standard deviation as surrogate for population value]. ii) One tailed test

The above problem may be solved for alternative hypothesis that<14% So, Ho : = 14%, X = 12 ; Alpha = 5% H1 : < 14% ; N = 100 ; = 4%

SE of mean = a / N = 4// 100 = 0.4 Z = (X - ) / SE of mean = (12.8 14)/0.4 = 3 The same as we got earlier. Difference in procedure arises hereafter only. The table value for 5% significance level for one-tailed test (onetailed test has alternative hypothesis of the form > or <, as against two-way test where alternative hypothesis is of the "Not equal to" form) is 1.645. As the calculated "Z" is greater than 1.645, the null hypothesis stands rejected. 2.4.2 Testing of difference between population means To test whether two populations have different means, a test procedure is followed. The same is dealt below, The Ho is taken as: 1 = 2 The H1 can be any of these three: (i) 1 2 ; (ii) 1 < 2 ; (iii) 1 > 2 The SE of difference between mean is given by: 1/n1 + 1/n2 where is population standard deviation, n1 = size of sample one and n2 = size of sample two. If is not available and sample standard deviations are available, the, SE is calculated as: S12 /n1 + S22/n2 The test statistic is (XI - X2)/SE, If this value is less than table value, Ho is accepted, otherwise rejected. In testing difference between means, if both nl and n2 are greater than 30, we use normal distribution, whether is known or only "s" values are known.

The mean Earnings per share of a sample of 50 southern textile mills is Rs. 8 while that of northern textile mills is Rs. 7, The for all textile mills is Rs. 2.6. Can we say the two regions differ in mean? Solution: Ho : 1 = 2 H1 : 2 2 Alpha = 5%; = 2.6 XI = 8; X2 = 7; n1 = n2 = 50
S.E. of diff. in means S1/n1 + 1/n2 2.6 1/50 + 1/50 = 0.52

Test Statistic = (XI - X2)/SE - (8-7)/.52 = 1.92 At 5% significance level the table value is 1.96. The calculated value is less than table value. So, Ho is accepted. That is, the mean EPS of southern mills is equal to mean EPS of northern mills. If we had adopted one tailed test with the H1: 1 > 2, the Ho will get rejected since, as calculated value > Critical value 1.645 at 5% level. 2.4.3 Testing for Significance of Standard Deviation The standard deviation of a population is assumed to be 5. A sample of 50 units produced a sample standard deviation of 4. Is there significant difference? We can test and tell. Ho : = 5 H1 : 5 S=4 = 5% The SE of SD is given by = S/2n = 4/2 x 50 = 4/100 = 0.4 Test static = (S - )/SE - (4-5)/0.4 = 2.5

The calculated value, viz. 2.5 is more than table value of Z, at 5% significance level, name 1.96. So, Ho is rejected. That is, the hypothesized value of standard deviation is rejected as it is not true. The value is not equal to 5. Testing for difference between standard deviations The significance of difference between two standard deviations can be tested given, sample sizes, sample standard deviations and table values. The S.E of diff. Between Std. Deviations = S12/2n1+S22/2n2 Test statistic = (s1 - s2)/SE The mean price of a scrip over a 100 trading days period and the standard iation were: Rs. 540 & Rs. 30. For another scrip, for an 80 days period the and standard deviation were Rs. 552 and Rs. 25. Test for equality of .dard deviation against, H1 : s1 s2 and ii) H1 : s1 > s2 Solution Set I Ho : 1 = 2 Ho : 1 2 S1 = 30; S2 = 25 N1 = 100 ; N2 = 80 Alpha = 5% Set II Ho : 1 = 2 Ho : 1 > 2 S1 = 30; S2 = 25 N1 = 100 ; N2 = 80 Alpha = 5%

S.E. = S12/2n1+S22/2n2 = 302/2 x 100 + 252/2x80 = 900/200 + 625/160 = 2.899 Test statistic = (S1 - S2)/SE - (30-25)/2.899 = 1.722

The table for 5% significance level for two tail test =1.96 i) ii) The calculated value is less than table value. So, null hypothesis is accepted. In the 2nd set of hypothesis, the null hypothesis is rejected, because calculated test statistic - 1.72, is greater than 'z' value for 5% significance level under one tailed test, which is equal to 1.645. IV. Testing for Variance Testing for variance, ie., Standard deviation squared, is done usmg chi-square distribution in the case of one population problems and Fdistribution for two population problems. The two problems already done using z-test, are now done using chi-square test and F-test as well. 2.4.4 Given : Ho : = 5 H1 : 5 S=4 = 5% n=
2

Testing for significance of variance : X2 Test

- (n-l)S2 / 2 = (50-1) x 42 / 52 = 31.36

The calculated value must fall between lower and upper x2 values obtained from table for given degrees of freedom and level of significance for accepting the Ho in the case two failed test. The critical X2 for 49 degrees of freedom and 5% level is 32.4 on right tail and 71.4 in the left tail. Since the calculated X2 is within the two critical values, Ho is accepted. So, the assumed population variance of 25 (i.e., 5x5) is the correct value.

For one tailed test with HI: a < 5, the critical x2 = 37.7 for 10% level and 49 degrees of freedom- The calculated value falls in the rejection region, being less than 37.7, the critical value. So, Ho is rejected. Testing for significance of difference in variances: F-test The problem on significance of difference between population variances is tested through 'F' test. We have to calculate F-ratio, given as S12/S22, where S12 is the larger variance of the given two sample variances and S22 is the other one. In our case F= S12/S22 = 302/252 = 900/625 = 1.44 For one tailed test, critical F value for 30-1= 29 degrees of freedom for numerator and 25-1 = 24 degree of freedom for denominator at 5% level is 1.94. The calculated value falls in the acceptance region. Ho is accepted. For two tailed test, we have to set lower and upper 'F' critical ratios. If the actual value falls within these, Ho is accepted. In our problem, the acceptance region extends from

F(29,24,0.95) to F(29,24,0.05). We can get F(29,24,0.05) directly from F table and is equal to 1.94. We cannot get F(29,24,0.95) directly from table. But that is equal to = 1/F(24,29,0.05) = 1/1.9 = 0.53

So the calculated 'F' value, 1.44 falls between the critical F values viz., 0.53 and 1.94. So, Ho is accepted. 2.4.5 Testing the significance of correlation coefficient large samples To test the significance of correlation coefficient, the following steps are involved: i) Calculate of standard error of correlation coefficient: The standard error of correlation coefficient is calculated as follows 1 r2 S.Er = N

ii)

Calculate of critical ratio Z The critical ratio Z is defined by the relation: r Z= S.Er

A random sample of 900 pairs from a bivariate normal population showed a correlation coefficient of +0.67. Is this value significant of correlation in the population? We set up the null hypothesis that the given correlation coefficient is not significant.

Ho: r is not significant H1: r is not significant

The standard error of correlation coefficient is given by

1 r2 S.Er = N =

1 0.672 = 0.018 900

The critical ratio Z is given by

r Z= S.Er =

0.67 = 37.22 0.018

At 5% level of significance, the rejection region consist of Z values equal to or less than -1.96 and equal to or greater than +1.96. The computed value of Z - 37.22 falls in the rejection region and the formulated null hypothesis is wrong.

2.4.6

Testing the significance of the difference between two independent correlation coefficients Suppose two independent samples of sizes and N2 give

NI

correlation coefficients r1 and r2 respectively. Here we have to test the hypothesis such that:

The samples are drawn from the same bivariate normal population. On the assumption that the samples are drawn from the same bivariate normal population, the difference between the values of Z for two samples is normally distributed with standard error:

S.E. Z1 Z2 =

1 N1 - 3

1 + N2 - 3

The test statistic 'Z' is defined by the relation: Z1 Z2 Z = S.E. Z1 Z2 And Z is normally distributed with mean zero and unit standard and deviation 1 + r1 Z1 = (1.1513) log10 1 - r1 and 1 + r2 Z2 = (1.1513) log10 1 r2

The computed values of Z is compared with table value/of Z at a specified level of significance and conclusion drawn. Illustration: The first of two samples consists of 23 pairs and gives a correlation coefficient of 0.5 while, the second of 28 pairs, has a correlation coefficient of 0.8. Are these values significantly different? Solution: We formulate the null hypothesis that these two correlation coefficients do not differ from each other: H0: , = 2; H1 : , 2

S.E. Z1 Z2 =

1 N1 - 3

1 + N2 - 3

1 23 - 3

1 + 28- 3

1 20

1 + 25 = 0.3

The statistic Z is defined by the relation:

Z1 Z2 Z = S.E. Z1 Z2 Where, 1 + r1 Z1 = (1.1513) log10 1 - r1 = 1.1513 log 3 - 1.1513 x 0.4771 =0.55 = (1.1513) log10 1 - 0.5 1 + 0.5

and

1 + r2 Z2 = (1.1513) log10 1 r2 = (1.1513) log10

1 + 0.8 1 - 0.8

= 1.1513 log 9 - 1.1513 x 0.9542 = 1.1 The statistic Z Z1 Z2 Z = S.E. Z1 Z2 Where Z1 = 0.55 ; Z2 = 1.1 and SEZI.Z2 = 0.3 Substituting these values :

0.55 1.1 Z = 0.3 = - 1.83

At 5% level of significance, the rejection region consists of Z values i.e. Z < -1.96 and Z > 1.96. The computed values of Z falls in the acceptance region. Hence the formulated null hypothesis is correct.

2.4.7

Testing of significance of population proportion Suppose that the promotable employees in an organisation is

40%. A random sample of employees from that population reveals that 50% are promotable or only 32% promotable. We now doubt whether the assumption of 40% employees as promotable is correct or not. In other words, whether the true population proportion of promotable employees is 40% or some other figure Similarly, supposes that the percent of defectives be 5% in a process. Whether truly the defectives percent is 5%, testing of significance of proportion can be done. i) i) ii) Testing for significance of population proportion Set the Ho : P = A specific value Set the Ho : P A specific value (or) P > A specific value (or) P < A specific value iii) Set the level of significance

Test Procedure

iv) Get the sample proportion v) Compute the standard error of proportion: the formula is P.Q. Sep = n Where, P is assumed population proportion, Q = 1 - P and n is the size of sample vi) vii) Compute the test statistic Compare the test statistic with the critical 'Z1 value. If the

calculated Z value is less than critical Z value, Ho is accepted. If calculated value is equal to or greater than critical Z, Ho is rejected. The HRD division head tells the MD of the company that 80% of staff are promotable, assess The MD constructed a committee to the-iromotability of employees. The committee drew a random

sample of 150 employees and found the promatability as 70%. In the circumstances, can we hold the population proportion of promotable employees at 80% as assumed earlier? We can take a decision through a statistical testing procedure. Let us test the H0 that P = 0.8 against H1: P 0.8 as well as P < 0.8 both at 5% level.

P.Q. Sep = n =

(8) (.2) = 150 0.0327

P-P Z = SEP =

0.7 0.8 = 0.0327 - 3.06

Since the calculated value of Z falls in the rejection region (-1.96 for two-mailed test), Ho is rejected. A Ho, rejected under two tailed test, will automatically get rejected under one way test as well. ii) Testing for significance of different between proportions An investment consultancy firm finds that 71% of 100 investors in city A prefer equity investment and 64.4% of 90 investors in city B prefer equity investment against debt investment. Whether the two cities differ in the proportion of investors preferring equity? Here, testing of significance of difference between population proportions has to be made to come to a conclusion. H0 : PA = PB (proportion of city A = proportion of city B) H1 : PA PB [two-tailed test] or _ PA > PB [one tailed test] PA = 0.71 qA = 0.29 = 5% PB = 0.644 PB = 0.356 nA = 100 ; nB = 90

SE PA - PB =

PcQc n1

PcQc n2

where Pc = combined proportion qc = 1 - combined proportion nA PA + nB PB Pc = nA + nB = (71.58)/190 = 0.6789 Qc = 1- 0.6789 = 0.3211 = 100 + 90 (100)(071) + (90)(.644)

SE PA - PB =

(0.6789)(0.3211) 100

(0.6789)(0.3211) 90

= 0.0678 Z = (PA-PB)/SE = (0.71 - 0.644) / 0.0678 = Q.973 The calculated Z, is less than 1.96, say Ho is accepted, under two tailed test. Also, as Z is less than 1.645, Ho is accepted under one tailed test as: well. 2.4.8 CM-Square Test (X2) The x2 test is one of the simplest and most widely used parametric as well as non-parametric tests in statistical work. The Chi-square test was firsl used by Karl Pearson.

Formula (O - E)2

X2

E Steps: 1. Calculate expected frequencies. In general the expected

frequency for any cell can be calculated from the following equation for contingency takes. For others as per given proportions or average of total the expect frequency is computed. RT X CT E
=

N E RT CT N 2. = = = = Expected frequency Row Total Column Total Total No. of observations

Take the difference between observed and expected frequencies and obtain the squares of these differences (O -E)

3.

Divide the values of (O -E)2 obtained in step (2) by the respective expected frequency and obtain the total [ ( O-E)2 /E ] and that total is the X2 value.

After ascertaining

x 2 value, the x 2 table is consulted to obtain

0.01

critical X2 value. The table comprises columns headed with symbols 0.05 for 5% level of significance and for 1% level of significance

while rows refer to degrees of freedom. If the calculated value falls in acceptance region, Ho is accepted and vice versa.

personnel

manager

is

trying

thedetermine

whether

absenteeism is greater on one day of the week than on another. His records for the past years show the sample. Days of a week No.of absentees Mon 66 Tue 57 Wed 54 48 Thu 75 Fri

Test whether the absenteeism is uniformly distributed. Solution: Let the Ho be; Absence is uniformly distributed over the week. On the basis of this hypothesis we should expect 60 absentees (66 + 57 + 48 + 75 = 300 / 5) on each day of a week. The chi-square is computed as follows : O 66 57 54 48 75 300 E 60 60 60 60 60 300 O-E +6 -3 -6 -12 15 0 (0-E)2 36 9 36 144 225 (O-E)2/E 0.60 0.15 0.60 2.40 3.75 7.50

X2 = (O - E)2/E = 7.50
Degrees of freedom = n 1 = 5-1 = 4

X2 0.05 = 9.49

The calculated value of X2 is less than the table value. The null hypothesis' is accepted. Hence we may conclude that the absence rate is spread uniformly over the week.

2.4.9 Non-parametric test ; Most of the statistical tests require an important assumption to be met if they are to be correctly applied. This assumption is that population from which samples are drawn is normally distributed. These statistical tests allow considerable latitude and deviation from normality. But there are situations when the researcher cannot or does not want to make such assumptions. In such situations we use statistical methods for testing hypothesis which are called "nonparametric" tests because such tests do not depend on any assumption about the parameters of the parent population. Some of the commonly used non-parametric tests are: 1. Sign test 2. Rank sum test 3. One sample run test 4. Kruskal-Wallis test 5. Spearman Rank Correlation test

The discussion of these is beyond the scope our study material. Interested learners may refer good text on this.

2.5

RESEARCH DESIGN A research design is nothing but a detailed plan of action for the

research. In Lesson I, under the heading research process, a series of steps in research is given. That is also a type of research design. A researcher attempting to solve the research problem, should necessarily prepare a plan which will help him to attain his ultimate motto, This plan is nothing but a research design. It is a plan which defines problem, identifies data needed, decides on tool of data collection, and type of study i.e., sample or case or census, picks up the sample or case or the population units as the case may be and goes about the collection and analysis of data. It is a tentative plan which .undergoes many modifications as the study progresses. It presents a series of guide posts to enable the researcher to progress in the right direction. According ; to Pauline V. Young, a research design is "the logical and systematic planning and directing a piece of research. The design, according to her "results from translating a general scientific model into varied research procedures". It gives an outline of the structure and process of the research programme. Without such a plan of study no scientific study is possible. 2.5.1 Definition Several definitions of 'research design' have been advanced by several writers on the subject of research methodology. Few of them are presented Blow: According to Claire Seltiz, Research Design is a catalogue of the various and facts relating to the formulation of a research effort. It is the arrangement of conditions for collection and analysis of data in a manner that aims to combine relevance to the research purpose with economy in procedure.

According to Paul E. Green and Tull, a research design is the specification of methods and procedures for acquiring the information needed. It is the over all operational pattern or framework, of the project that stipulates what information is to be collected from which sources by what procedures. If it is a good design, it will ensure that the information obtained is relevant to the j research questions and that it was collected by objective and economical procedures. According to Bernard S. Phillips, the research design constitutes the blueprint for the collection, measurement and analysis of data. It aids the scientist in the allocation of this limited resources by posing crucial choices: Is the blueprint to include experiments,. interviews, observation, the analysis of records, simulation or some combination of these? Are the methods of data collection and research situation to be highly structured? Is an intensive study of a small sample more effective than a less intensive study of a large sample? Should the analysis be primarily quantitative or qualitative. According to Fred N. Kerlingar, research design is the plan, structure and strategy of investigation conceived so as to obtain answers to research questions and to control variance. The plan is the overall scheme or programme of research. It includes an outline of what the investigator will do from writing the hypotheses and their operational implications to the final analysis of the data. To structure the research is to outline the scheme and paradigm of the operations of the variables strategy. It includes the methods to be used to gather and analyse the data. In other words, strategy implies how the research objectives will be reached and how the problems encountered in the research will be tackled.

2.5.2

Factors Affecting Research Design

The following factors affect the design of research: i) ii) iii) iv) v) vi) vii) viii) the nature of the studs the scope of the study the level of the study availability of sufficient data; proper exposure to the sources of data, especially primary data; availability of time; availability of money and manpower; impact of the various internal and external as well as controllable and uncontrollable variables on the research project; ix) x) the ability, skill, knowledge, and technical background of the researcher; and utility and applicability of the research result in practice.

2.5.3

Aspects of Research Design Research design addresses the variables (dependent,

independent and extraneous), relationship among variables and control thereof, hypothesis, design of experiment, conduct and control of the experiment and so on. Research design can be dissected and put into the following: Problem design, Objectives design, Hypothesis design, Tool design, Sampling design, Experiment design, Survey/field design, Statistical design, Analytical design, Interpretation design and Report design. These are explained at different places in the study material and hence not dealt here.

2.5.4

Evaluation of Research Design The research design must be good. The question of good design

is related to the purpose or objective of the research problem and also with the nature of it the problem to be studied. A good design is often characterised by features like flexibility, appropriateness, economical and so on. Generally, the design which minimizes bias and maximises the reliability of the data collected and analysed is considered a good design. The design which gives the smallest experimental error is supposed to be the best design in many investigations. Similarly, a design which yields maximal information and provides an opportunity for considering many different aspects of a problem is considered most appropriate and efficient design in respect of many research problems. A design may be quite suitable in one case, but may be found wanting in one respect or the other in the context of some other research problem. The fundamental questions in evaluating a research design pertain to the precision, reliability and relevance of the data and their analysis. Before actually carrying out research, it is better if the researcher revaluates his research design. This can be achieved if he verifies the following aspects for their explicitness. i) ii) iii) iv) How relevant are the objectives? How relevant are the hypotheses? How explicit are the hypotheses? Have the problems and hypotheses been stated in

operational terms scientifically?

v)

Has the plan of research been presented in detail so that its logic is apparent?

vi) vii) viii) ix) x)

How scientific is data collection tool? How scientific is method of data collection? How precise are the observations? Can other investigators repeat the observations? Do the data actually satisfy the demands of the problem, i.e. do they actually demonstrate the conclusion?

xi)

Does the research design ensure a comparison that is not subject to the alternate interpretations?

xii)

Are the statistical designs appropriate?

QUESTIONS 1. 2. 3. 4. How do you plan your research Explain the meaning and issues in problem formulation A problem will put is half-solved" - commit Explain the meaning, need and significance of review of literature 5. 6. 7. 8. 9. Explain the meaning, source and types of hypothesis. Explain procedure for hypotheses testing Explain only two parameters tests of hypotheses. What is research design? State the need for the same. State the factors that affect research design.

10. How do you evaluate research designs? 11. Explain the procedure of Hypothesis testing. 12. What do you mean by testing for significance of mean and difference between means? Explain with an example.

13. What do you mean by testing for significance of proportion and difference between proportions? Explain with an example. 14. What do you mean by testing for significance of variance and difference between variables? Explain with an example. 15. What do you mean by testing for significance of std, deviation and difference between std. deviations? Explain with an example. 16. What do you mean by testing for significance of correlation and difference between correlation coefficients? Explain with an example. 17. The assumed population mean is 75. A sample of size 100 gives a mean of. 78 and std. deviation 2. Test whether the population mean has increased ? 18. The assumed population proportion is 0.8. A sample of 25 gives a proportion of 0.72. Test at 5% level whether population proportion has decreased. 19. Two samples drawn from two normal populations have means of 25 and 29, the sample sizes being 42 and 38. The std. deviation of the samples is 3 and 3.5. Test whether two populations have identical means. 20. Two samples drawn from two normal populations have proportions of 25 and 29, the sample sizes being 42 and 38. Test whether 2 populations have identical proportions. 21. Two samples drawn from two normal populations have Test variances of 25 36, the sample sizes being 42 and 38. whether the two dilations have identical variances. 22. Two samples drawn from two normal populations of bivariate relationship have correlation coefficient of 0.8 and 0.7, the sample sizes being 42 and 38. Test whether 2 populations have identical correlation coefficients. Use Z transformation model.

REFERENCES 1. 2. 3. 4. 5. 6. 7. Research Materials - P.Saravanavel Elements of Educational Research - S.P.Sukhia Research for Marketing Decisions - Green & Tull Marketing Research- Sangeetha Agarwal Research Methodology - Basetia and Sharma Encyclopedia of Modern Education Research - S.K. Sharma Research Methodology - C.R. Kothari

********

UNIT-III SAMPLING DESIGN

In this lesson, the concepts of sample, census, sampling design, merits and demerits of sampling method, essentials of a good sampling method methods of sampling, procedure for selecting a sample, determining the size of sample and sampling and non-sampling errors are dealt. Introduction Research involves gathering information pertaining to the objectives the research study from relevant people or entities studied. Now a question arises. How many people or entitles of relevant population be contacted. All or a representative group of people be contacted? Similarly, should all the entities a representative or a section of the entities be covered? If it is decided to study/contact all of the people or entities such course is called census study and if only a section or group considered representative of the whole population is covered that is called sample study. 3.1 DEFINITIONS We are concerned with certain terms here. These are

population elements, sample, sample frame, etc. i) Population Population means the aggregate of all the persons, elements, objects, or items having defined feature(s) such as occupying certain geographical space or having a certain quality / qualities or having a common origin or background or such other attribute or feature.

In a study on user satisfaction of internet services, to define the population first we have to curve out the geographical area. Say a particular telecom circle. People in that telecom circle using internet services at home or office or at internet cafe will form the population. There are subcategories in the population, namely, users at homes, users at offices and users of internet at internet cafes. There are men and women users. Internet can be used for many purposes: Sending/receiving mails, browsing and downloading, on-line ordering etc. Accordingly, subgroups may be carried out based on services obtained. By combining different bases of classification compound sub-categories can be effected. Suppose in a given telecom circle in a given period, say, Jan -December in a particular year, 36256 internet users were recorded. They constitute the population. Of them, 12984 are say women and be rest 23272 are men. Thus, the gender base classification of population splits the population into, two exclusive sub-groups, exclusive groups, because a person may be a male or female. A person may use it to transmit mails or browse and download or place orders. Accordingly sub-groups by. use arise. But exclusive sub-groups do not arise, because a person may use internet for more than one purpose. Defining a population and sub-grouping the same are really tough intellectual aspects. ii) Elements of Population Elements of a population mean the resources or items or objects forming a population. In the above internet users example, the elements are the individual users. Men users and women users give gender-split population elements, E-main users and browsers give purpose-based population elements. And so on. Simple elements, ie; no classification into sub-categories and compound elements, thus, are two classes of population elements.

The elements of a population must conform to chosen specifications such as space, time, attribute and variables. Space refers to a geographical expanse or coverage. Time refers to a particular moment or a period of time. Attribute refers to character possessed, (i) male and female, (ii) white, brown or black and so on Variable is the feature - on which sample elements vary from one another apart from the, attribute factor. Income vary, education levels vary, etc. These are variables. iii) Census In the internet users example if you study the whole of 36256 internet users in the given telecom circle, that becomes a census study. It is total enumeration of all individual elements or units of the population or universe. iv) Sample Chambers 20th Century Dictionary defines a sample as a specimen or a small portion to show the quality of the whole. In the internet users example in the given telecom circle, you may study only, say 10% or 15% or 30% of the 36256 users. Then those of the users who you contact constitute the sample. You may decide to contact 10% of women and 8% of men users. Then these two sub-samples make your sample. v) Sample Frame Sample frame enumerates and enlists the population in terms of elements, simple or compound as the case may be, as decided to be used for conducting the survey. The sample frame constitutes the source list from which samples are drawing. A well defined mutually

exclusive and collectively exhaustive sample frame is the hall mark of quality research. But it is difficult to create one. vi) Stratification Stratification is dividing the population into sub-segments based on chosen attributes or variables range. vii) Sampling Design Sampling design is description of basis on which sample is drawn from the population. Viii) Statistic The mean, mode, correlation, or any statistical property estimated or Computed for a sample is called statistic. ix) Parameters The mean, mode, correlation or any statistical property estimated or computed for a population is called parameter. Mean income, mean productivity std. deviation of mileage, etc. are parametric in character. That is a population value exists, whereas mean preference of people for a brand or mean association between education and income does not exist. These are non-parametric in nature. Later in lesson 4, we talk about parametric and nonparametric tests. Forget not to recollect the meaning of these terms. x) Population Vs Universe Universe is broad enough to cover all elements that theoretically meet the attributes - set so as to be called as belonging to a group.

Population

is

said

to

cover

all

elements

that

practically

or

operationally meet the attributes - set. Suppose customer satisfaction in respect of telephone services is your study. Everyone, who uses telephone services constitutes the universe, theoretically. To practically execute the study, an element may be defined as one who owns a telephone connection as on date and has got the connection at least one year ago and whose bimonthly average telephone bill in the recently concluded year is at] least Rs.1000 or so. These qualifications give definite shape to population. While universe is infinite, a population as defined above is a finite one. An infinite universe is like the drops in an ocean, the starts in the sky, the executive styles or creative talents or objective type questions on current affairs. A finite population is like the one given above. We can give other examples: those doing their MBA degree in Alagappa University during a particular year; the registered owners of TVS scooty or Maruti Zen VX and the like. 3.2 CENSUS METHOD Census involves complete enumeration of all elements of the population Suppose you want to study the passenger car industry. In a country hardly there may be five or six companies manufacturing cars, Similarly bikes, scooters planes, ships, buses, train coaches, bicycles, televisions, telephone handsets, and so on. Each producer company may be studied thus ensuring a complete enumeration of all. Similarly census method is needed in examinations. Every candidates answer paper has to be valued thoroughly. In elections, cent percent counting of polled votes is needed. Of course, total enumeration of whole of population at least once in 10 years is needed to know the population trend in a country. Similarly enumeration of

small scale, medium scale and large scale business units is done to have the aggregate picture of industrial units. 3.2.1 Use of Census method of Investigation In business research census method is adopted in certain cases: complete study of all firms in an industry, especially one with few dominant firms like the case iron and steel industry, automobile industry, POL industry etc., ii) complete study, of concessions such as duty drawback, export market development assistance, availed by industries, iii) complete study of portion participating in a trade fair like the one organised by India Trade Promotion Organisation, New Delhi and iv) thorough study of mutual funds in a country during a certain year. 3.2.2 Merits of Census method Census method is ensuring 100% accuracy and 100% coverage. Minute details of the studied population is fully known. There is no sampling error or error due to bias in selection of samples. The characters of the population are maintained in original form, ie., no extrapolation, no intrapolation or no guesstimates or estimates. Purity is thus well maintained. It is more reliable. No need for statistical testing or estimation. 3.2.3 Demerits of Census method Cost escalating, time consuming. While sample error may be done away with, other errors in any investigation are not ruled out. Hence the so called 100% accuracy is full of *ifs' and 'buts'. Even assuming high accuracy, is that worth the high cost associated with it? No. In certain cases census means destroying the whole population like testing the longevity of bulbs which means, continuous burning of

bulbs till they go fuse. Further, it is silly to go for census when you test the taste of something. You need not eat whole thing to test the taste or whether sufficient cooking has been achieved. They chef does taste a little and squeeze of pinch to test the extent of cooking. One can't eat everything nor pinch every bit. So, census method lacks practicality. Hence the preference f for sampling. 3.3 SAMPLING METHOD Almost all research works are based on ________________. Instead of analysing a heap of census data unmethodically, systematic analysis of data pertaining to a sample facilitates a more complete understanding of the universe. In this section the varied aspects of sampling are studied. Sampling means taking out a sample for the purpose of the study. Sampling is a procedural pursuit to be followed in picking out the sample elements. Against a census study, a sample study has many merits. Of course, there are demerits too. 3.3.1 Merits of Sampling Sampling is used because it offers many advantages over taking a census. First, it saves on inputs. By interviewing a hundred customers in a town rather than lakh living there, we certainly save a lot of money, time and energy too. Time in many cases, is as important as money, as most research projects are time-bound. Second, though it may sound surprising, a sample may be more accurate than a census. The difference between a true figure like the average income of d population and the figure that is actually

obtained from our research based on a sample is called error. The error may arise because of the sampling itself, since a sample may have a different average from that of the population. The difference is called 'sampling error. Sometimes the difference arises because of inefficiency in collection, editing and processing of data; then, it is called 'non-sampling error. In the case of a well-taken sample, the science of statistics helps us to tell accurately the sampling error, that is, the range within which the population mean would lie on either side of the sample mean obtained. But in the case of a census, more staff have to be appointed and hence the quality of the enumerators and other staff would be lower, their efficiency in collection of data, editing and processing would be poorer and there would be greater chances of non-sampling errors than in a sample. Third, sampling is inevitable when the study results in destruction of the sample element. A company making electric lamps cannot test every one of them with respect to its maximum life or voltage resistance because then would have nothing left to sell. Fourth, sampling is highly practical and intensive. Intensive analysis is possible with sample survey. Sampling is good enough to have full understanding in most cases. 3.3.2 Demerits of Sampling Sampling has certain demerits, of course. First, it tasks accuracy, because all elements are not included. The excluded elements may be theme cases and that their exclusion gives erroneous conclusions. Second, it is difficult to response from all sample units and that deviation from selected sample needed. This leads to findings of the sample study get changed. Third, size of I sample remaining more a mark of opinion diverse views prevails. Fox sampling errors occur and expert knowledge is needed to keep the errors under check.

Fifth, biased sample, unrepresentative sample, problems. 3.4 PRINCIPLES OF SAMPLING

and the like cat

Sampling is based on certain mathematical and statistical principles. These are: i) Principle of statistical regularity, ii) Principle of inertia of large numbers, iii) Central Limit theorem and iv) Symmetrical Frequency distribution. 3.4.1 Principle of Statistical Regularity The principle of statistical regularity states that when a random and representative sample is taken from a population the sample is very much likely to possess almost the same characteristics as that of the population. This is' so, because the variability in the features of elementary units of a population has limits. Mangoes of a particular variety vary in weight, shape, size, colour and sweetness, but the variations are within limits and that no mango can smell like jackfruit, taste like watermelon, figure like a banana or is sized like a pineapple. 3.4.2 Principle of Inertia of Large Numbers The principle of Inertia of large numbers is a corollary of the principle of statistical regularity. The inertia principle states that as the size of sample rises the probability that the sample statistic is close population parameter moves close to l. In other words, larger the size of sample, smaller the variation between values of sample statistics and relevant population parameters.

3.4.3

Central limit theorem The central limit theorem tells that the sampling distribution of

any statistic is normal irrespective of the shape of the population distribution from which the samples are drawn. Therefore inferences about population can be le from features are the samples drawn from the same. 3.4.4 Symmetrical Distribution Sampling is used because it offers many advantages over census. First, it save on inputs. By interviewing a hundred customers in a town rather than the living there, we certainly save a lot of money, time and energy too. Time in many cases, is as important as money, as most research projects are time-bound. Second, though it may sound surprising, a sample may be more accurate a census. The difference between a true figure like the average income of a population and the figure that is actually obtained from our research based on a sample is called error. The error may arise because of the sampling itself, since a sample may have a different average from that of the population. The difference is called sampling error. Sometimes the difference arises because of inefficiency in collection, editing and processing of data; then, it is called 'non-sampling error'. In the case of a well-taken sample, the science of statistics helps us to tell accurately the sampling error, that is, the range within which the population mean would lie on either side of the sample mean obtained. But in the case of a census, more staff have to be appointed and hence the quality of the enumerators and other staff would be lower, their efficiency in collection of data, editing and processing would be poorer and there would be greater chances of non-sampling errors than in a sample.

Third, sampling is inevitable when the study results in testing each sample element. A company making electric lamps- cannot test every one of them with respect to its maximum life or voltage resistance because then would have nothing left to sell. Fourth, sampling is highly practical and intensive. Intensive analysis possible with sample survey. Sampling is good enough to have understanding in most cases. Finally, a survey focuses to get knowledge of the population. This is by studying a representative portion of the population. No need for studying units exists. 3.5 ESSENTIALS OF A GOOD SAMPLE A good sample must satisfy certain requisites. These

are representativeness, ii) appropriate size, iii) goal oriented, iv) measurability, practicability, vi) economical. These essential requisites of a good sample, dealt below: 3.5.1 Representativeness Representativeness of the population by the sample must be ensured before further pursuit is made in the course of research. The sample must representative of the population. Usually any random sample, because operation of the principles of statistical regularity and inertia of large numbers representative of the population. 3.5.2 Appropriate size The sample size should be appropriate neither too large nor too Larger the sample higher the accuracy and vice versa. But mere size do ensure accuracy. Appropriate stratification, due proportioning of

sample different strata and such other scientific methods ensure accuracy. It is be least the sample is statistically larger, ie equal to or greater than 30. 3.5.3 Goal Orientation The sample must reflect the goal of the research. If the researcher' is just formulation of hypothesis, a less exact sampling exercise may be enough. If the goal is testing of hypothesis a higher exactness is needed in the san exercise. If the purpose is estimating population parameter, still higher exactness is needed in the scheme of sampling. 3.5.4 Measurability The sample must help measuring sampling error from sample result: know sampling error of mean is = pq/n; sampling error of proportion is: and so on, where, is population std. deviation, p = population success proportion, n = sample size. To ensure measurability of sampling error, random sampling need to be adopted. 3.5.5 Practicability A sample must be capable of being practised. The size, the composition of sample elements, the location of the sample elements, etc. must pi contacting them and getting the responses. About 2.5 mn foreign tourists visited India in 2001. Even 1% of facts tourists means a size of 25000 tourists, possible to meet and collect data from this large number of the sample units. Most of them had already returned home. How can we elicit views from tf , These are the impracticability a sample design must avoid.

3.5.6

Economical One of the main reasons for use of sample studies is the

economic; inexpensiveness of this method. So, a sample must be economical to the research firm for contact and collection. 3.6 METHODS OF SAMPLING Let us now look at the methods that can be adopted for selecting the units from the sample frame. The methods of sampling can be broadly classified into i) probability sampling and ii) nonprobability sampling, as follows:

Sampling Methods

Probability methods 1. 2. Simple random sample Stratified sample a) Proportionate b) Disproportionate 3. Cluster Sample a) Systematic sample b) Area sample 1. 2. 3. 4.

Non-Probability methods Judgment Sample Snowball sample Convenience sample Quota sample

An explanation of each these methods is given below. 3.6.1 Probability Sampling Methods Probability sampling ensures that each element of the

population stratum has equal probability of being selected. Probability sampling methods are described below:

3.6.1.1

Simple Random Sampling

In probability sampling, each element of the population has a known non-zero chance of being selected for the sample. Among the probability sampling methods, simple random sampling is simplest as its name indicate and it underlies many of the more complex methods. However, it is not a widely used method. Here, every element of the population has an equal chance of getting selected. The following^ is a description of one of the ways a simple random sample can be selected, namely, the lottery method. Lottery method is very simple and easy to follow where the population is not too large. Suppose twenty employees are to be chosen from a thousand employees of a factory for a study of job satisfaction. The name of each employee is written on a piece of paper of uniform size. The pieces of paper are thoroughly mixed before and after being placed on a large bowl. Then twenty pieces are drawn out without looking at the name. The sample so selected would be a simple random sample. Here each employee in the factory has twenty chances of being selected. It should be understood that random sampling is a highly scientific, equal probability/selection procedure, and it is not the haphazard sampling method of meeting whoever crosses your way, though the dictionary may define the word random' as haphazard. The haphazard 'man in the street interview' does not ^require any sample frame or scientific selection procedure and it is a non-probability method of sampling called convenience sampling or accidental sampling and described later in this lesson. The lottery method may also be ^adopted using tokens numbering 0 to 9 in a container. Picking them up one after pother after replacement we can get random numbers. For picking the twenty employees first, the names of the thousand employees should be\ alphabetically arranged and serially numbered, using the tokens we have to get three digit random numbers. Suppose we pick 3 first, 0 next, and 8 next. The employees bearing the number

308 gets selected for the sample. If 000 is the next number, person bearing the number 1000 gets selected, ie. 001 twenty persons selected. The lottery procedure, however, is tedious. So random number tables are for selecting a simple random sample. For this we should take a book of numbers, choose a page in it at random, and start from a line or column page at random. Suppose the numbers we get read 3125 5750 8144 1496 9867 5454 4905 4099 6703 9967 2082 3074 5414 7884 6836 is selected the employee whose name if first in the list is selected and so on, till

Then, the twenty employees bearing the following numbers respectively are chosen from the sample frame: 312 992 514 82 964 788 905 481 996 445 754 454 145 670 750 330 986 746 740 836

If the same number appears again, it is dropped and the next number taken. How these numbers are got? From the first random number 312 is taken. The left out number is 5. This is prefixed to the first two numbers of the next random number, 1496 and we get 514. Then with the left out 96, first digit of the next random number 4905, ie., 4 is suffixed. And we get 964 and so on. Now computers are - available and they have programmes to general random numbers to meet our needs.

3.6.1.2

Stratified Random Sampling

In the above example of choosing a simple random sample of twenty employees out of a thousand in a factory, suppose they include 100 supervise and 900 workers. A simple random sample comes by mere chance and it possible that among the twenty chosen, all may be supervisors or none may be. If the researcher feels that he should study both the subgroups, it would be me sensible to take a random sample from each subgroup (stratum) after separate lists for the two strata. Such sampling be is called in stratified rand sampling. on differ Stratification may done business research

characteristics like sex, age (e.g. young, middle-aged, old), race, religion occupation, education, residential area (e.g., rural, urban) ownership of (e.g., public sector, private sector), size of business or income. The satisfaction may be proportionate or disproportionate. In the
CE

sampling of twenty employees from a factory described above, if

2 supervisors and 18 workers are chosen at random to form the sample of twenty, the sample division between supervisor and worker is in the same proportion as ml population (1:9), and it is a stratified proportionate sample. In this case worker and supervisors have the same sampling fraction, 1/50 which is the probability of each employee getting selected into the sample. Suppose on the other hand, 5 supervisors and 15 workers are chosen random from the 100 and 900 respectively. It is a stratified disproportionate sample. Here the sampling fraction is different for the two groups. In where you expect workers to have more homogeneity and the supervise have more variability in the relevant characteristics, a larger than proportionate sample of supervisors would be advisable. The rule here is, "the larger the variation, the larger the sample and vice versa.

In the case of a stratified sample as above, the supervisors and the workers do not have an equal chance of getting selected, but nevertheless, each employee has a known, non-zero chance and hence stratified sampling is also a probability sampling. We have already noted a reason why disproportionate

stratification may be adopted. There are also other reasons for this as well. Suppose the cost of collecting data from one stratum (say, rural population) is higher, its proportion may be reduced, and strata with lower data collection costs may be sampled at higher rates. Sometimes, the researcher is interested more in making comparisons between each stratum estimate rather than in aggregating them all into an overall population estimate. Then, we may take an equal number of items from the strata (e.g. 10 supervisors and 10 workers) assuming the variability of characteristics and cost of collection to be constant between the strata. However, proportionate stratification is used more than

disproportionate stratification because estimation is simpler and because it guarantees that the t estimates of population parameters is no less precise than those obtained from a ; simple random sample of the same size. In fact, with regard to accuracy of estimates, proportionate stratification is preferable to simple random sampling. But the problem here is that in addition to the overall sampling frame we require information on about which item belongs to which subgroup. We may, for example, are a sample frame from the voters list in a city for a study of housewives respect to, say, their purchase of certain goods; but if we wish to have the sample stratified on the basis of education, or between mothers and childless housewives, we require additional information about the educational level of the housewives and the number of children they have, to determine how many rives in each stratum we wish to interview.

3.6.1.3

Cluster Sampling

In cluster sampling the population again is subdivided into subgroups termed clusters instead of strata. The term cluster means a bunch of similar things. Suppose the thousand employees of a factory in our example comprise Steams each consisting of one supervisor and nine workers, and we choose "these teams at random, it is a cluster sample, because we have taken two; hundred clusters as our sample. 3.6.1.4 Systematic Sample

One popular type of cluster sample is the systematic random sample also called quasi-random sample. Suppose the thousand employees of the factory are listed alphabetically and numbered serially to form the sample frame. (The alphabetical order is usually a random arrangement with respect to most characteristics and so forms an unbiased sample frame). As twenty employees are to be chosen from the thousand, it is a one in fifty sample. So, here, 50 is the sampling interval or 'skip factor'. Now choose a number within fifty at random. Suppose you get 32. This is called the sampling start. Now choose every 50th number from the sampling start. Then we get the following twenty numbers for the sample: 32, 82, ... 932, 982. (in other words, if N/n = k, choose every kth number starting from a randomly chosen number within k for a systematic sample). In fact, the thousand employees in our example can be said to form 50 clusters on the basis of their numbers in the sample frame, and you have chosen one of the fifty at random. If a proportionate representation of supervisors and workers is required here, the sample frame may be arranged with the names of the hundred supervisors first, followed by those of the nine hundred worker, each it alphabetical order. If the sampled numbers are as above, then there would be of the supervisors selected (32 and 82) and eighteen of the workers (132, 182, 982) among the twenty chosen. In

the case of an ordered sample frame as above a systematic sample will prove to be statistically more efficient than a simple random sample. But there is one major problem with systematic sampling. If the sample frame has a cyclical pattern that coincides with a multiple of the size of sampling interval or vice versa, the result would be biased. Suppose, for a study of absenteeism, two days in the week are chosen from the five working days the week. If the two happen to be Monday and Friday, the average percentage^ absenteeism calculated from this study will be much higher than the actual daily average for the year because more people generally absent themselves on the two days since they are adjoint to the weekend holidays. Again, if the sample frame is a list of members of a social club, where the husband's name is follow by the wife's name such that every odd number is a male and every even number is a woman, a systematic sample of an even number is a woman, a systematic sample of an even number is a woman, a systematic sample of an even number series like 32, 82, 132, ... will represent women only, To give another example, if the selection interval is 10 and the sample start is also 10 in a marketing research of households; it is possible that every tenth house in a colony is a corner house, and corner houses are inhabited by people with higher than average income as they command a higher cent. In all the above cases, it is obvious that the sample survey will yield biased results. Another problem in systematic sampling is what to do when the sampling interval k is a fraction. Suppose five persons are to be selected from 32 by systematic sampling. The sampling interval is 32/5-6.4. In such cases select a number at random between 1 and 64. Suppose you get 8. You start with 0.8 and add up 6.4 repeatedly. You then get the series 0.8, 7.2, 13.6, 20.0, 26.4. Rounding them up you get the following numbers for the sample: 1, 7, 14, 20 and 26. The interval here is sometimes 6 and sometimes 7.

36.1.5

Area Sampling type of cluster sampling particularly useful in

Another

marketing research is area sampling. A town may be divided into a number of areas usually by placing a numbered grid on the map of the town. A sample of the areas is chosen at random and every household in the selected areas is studied. Hence the method is also known as grid sampling. The problem here is that households next to each other may not reflect the heterogeneity in the population. But the cost of collecting data will be quite low because there is less need for travel by the interviewer. 1.6.1.6 Multi-stage Sampling

One of the main problems in simple random sampling relates to the availability of the sample frame. As already stated, area sampling solves the problem to a certain extent, Here, the researcher needs to develop only the lists ^elected areas as clusters to choose from. In the selected clusters he would do 'Census study. But if he wants to have a sample at this stage too, he has to develop the list of population elements in the selected clusters before choosing a develop from each. In this case he conducts sampling of households in more than stage, that is, sampling clusters or areas and sampling of households in each. The National Sample Survey and the crop - cutting experiments foisting agricultural yields in India use multistage sampling. For example, a survey of villages with respect to some problem of national importance may be done as follows. Two states may be chosen at random first. Two district may then be chosen at random from each of the two states. Then, ten villages are selected at random from each of the districts. This yields a sample of 40 villages. At rural survey level too, there can be sampling of households. In the crop-cutting experiments, the villages are selected first, then fields in the villages and then randomly located cuts in the selected fields; the crop cut in

them and from the data gathered a forecast is made of the country's gain output. 3.6.1.7 Multiphase Sampling There are cases where a researcher adopts a multiphase sampling method too. Here a sample is taken for the first phase of research from which some simple and general information is gathered. Then, a subsample from this large sample is taken for a detailed enquiry in the second phase of the study. Suppose; you study the performance of Tamil Nadu candidates in the UPSC Civil Services: Exam. Your first sample is those who appear for the preliminary exam. The next is from this, who sit for the main exam. Next is from this who appear for the personality test. 3.6.2 Non-probability Sampling The need for and methods of non-probability sampling are dealt below. Need for Non-Probability Sampling Any of the following conditions may make a researcher adopt a nor probability method of sampling 1) Probability sampling is not practical for: reason like non-availability of sample frame, 2) Greater representativeness of 1 sample can be obtained with a non-probability method like judgment sample because sample size is very small. 3) Time is very short or a budget const exists. 4) Only rough approximations of population parameters are required, The study is an exploratory or pilot study to suggest hypotheses for testing regular study later.

3.6.2.1 Judgment Sampling In Judgment Sampling method which is also called purposive deliberate sampling the researcher uses his experience in selecting units that representative of the target population or that are most relevant to the issue study. Hindustan Lever once conducted a market research on how farmers use their additional income that resulted from the Green Revolution. For this, they deliberately chose four districts of India including Thanjavur and Ludhiana as representative areas which had gained best from the Green Revolution. If the sample size is to be very small, say, 10 or under, a judgmental sample is usually more reliable and representative of the population than a probability sample. Taking another example, suppose you want to study the problem faced by a certain small industry with respect to its export. There are 500 firms in the industry of which five are noted for being successful in the export sphere. If you wish to know the problems in exporting and how they can be solved best,, you would select only these five and not a random sample of any five out of the five hundred. You may however select a few in addition from the unsuccessful 495 if you wish to find out what prevented them from being successful. Judgment sampling is moderately used in practice. The consumer price index number is an example. Here, the cities covered and the retail units in the cities from which price data are collected are all judgment samples. Yet most of
IE

would believe that the index

number indicates price changes correctly for any place in the country. 3.6.2.2 Snowball sampling Again, suppose a focus group interview of eight or nine people with respect to, say, service quality of nationalised banks is to be made, a non-liability sample would be quite appropriate. One type of such sample is called 'snowball design' as mentioned by Aaker and

Day in their Marketing Research (John Wiley, 1983, p.312). Suppose you are interested in a marketing research on a new type of welding process but have no idea of which technologists in the country can express an opinion on it. You start with viewing Kumar of Pieco Ltd., then ask him to suggest another technologist who had used this process and can express an idea on it. He suggests Naidu of and Toubro. And from Naidu you go to someone whom he suggests. And if so on with each respondent identifying other possible respondents, the sample is building up like a snowball. Convenience Sampling If you stand outside a supermarket and ask customers coming out who are willing to answer questions relating to your study, you are taking what is called a convenience sample. It is also called haphazard or accidental sample, the latter name because the units enter into your sample by accident as they happened to be there at the time you were there. As the sample may not be representative of the population and there is no way to determine the sample error, this method is called sometimes derisively as a "quick and dirty study, Here, the researcher does not chose units scientifically from a sample frame but lakes whatever is convenient for him. Nevertheless, this method is extensively used in research when hypotheses are formed for less rigorous testing or where inaccuracies in estimation are not important but the study is required to be made quickly and cheaply. Opinion polls on matters like "which party would win the coming elections", "who should be the next President?" or "Should public sector industries be privatised?" are mostly based on convenience samples. The assumption who in convenience sampling is that those

are interviewed are no different from others belonging to the

target population not interviewed. Critics of this method question this

assumption. As there is considerably selection bias here, it is dangerous to attempt the use of the results, of a convenience sample to make inference about the general population. To quote an example, a TV station in USA asked citizens to phone the station up their opinion whether the age of drinking should be reduced from 21 to 18. 40C calls came in support of the reduction and very few were against. Those under 21 who were interested in the age reduction probably phoned up in support the elders who were against would not have cared to phone up. Can it be understood now ^hat the city's population is predominantly in favour of age reduction? 3.6.2.4 Quota Sampling The quota sampling technique has many variants and it is widely used market researchers in particular. It has all the advantages of convent sampling like cheapness, quickness, ease in administration and nonrequirement of sample frame, and in addition it is possible here to select persons possessing certain characteristics in proportion to their appearance in the total population. Here the interviewer is assigned certain quotas. For example, he may have, interview six men under 35 years of age, five men 35 or older, five employed women and eight unemployed women. He is free to choose any person satisfying these characteristics. This resembles stratified sampling, but it is not probability sample. Whereas in probability sampling there is the risk of non-response from a selected person, such a risk is absent in quota sampling interviewer can choose another person with the same characteristics if the first one does not respond to the interview. 3.6.3 Comparison Between Probability and Non-Probability

Sampling The relative merits of probability and non-probability sampling can be summarized on the basis of the following factors.

i)

Reliability Since sampling theory is based on chance selection, its

principles are not applicable to non-probability samples. In the case of non-probability we cannot get a reliable estimate of a universe characteristic from the sample data. Confidence limits do not apply to these samples. The reliability of the sample statistics cannot be measured. ii) Sizes If the size of the sample would be small because of, say, time consideration, it is better to use purposive sampling than to use random sampling. In probability sampling, precision of estimates increases only when the size of the sample is larger. There is no formula for determining optimum for non-probability samples. iii) Selection of Units In many studies, if you have adequate information about the universe and characteristics, judgment sampling would be effective. If you do not have the expertise to select typical units for the study, probability sampling should be better. iv) Cost and Time Non-probability sampling usually requires less money and less time. But if quality of information is more important than money cost and time, probability sampling is better.

3.7

SELECTION OF A SAMPLE Having decided the to go for sampling rather than census and

the method of sampling the next question is taking the sample. In this context a sampling frame must be developed. The size of sample must be decided. By applying the decided method of sampling, the sample is selected. 3.7.1 Sampling frame The sampling frame is the list of sampling units. Suppose your study is on. Determinants of Executive Excellence in Private Sector in India. You hence to define who is an executive and 'what constitutes India's private corporate sector. You can prepare a list of corporate sector undertakings in Indian private sector as you have defined. You have to then prepare a list of executives working in these undertakings. The list of executives may have different strata, if decided to have like that. This is the sampling frame. This is the list from which the sample is going to be drawn. The list must be mutually exclusive ( ie; an executive is included only once in the list) and collectively, exhaustive (ie; all serving executives are included). 3.7.2 Size of Sample You opt for sample survey. One immediate question is size of sample of What should be the size of sample? This is a nagging question to answer. In this regard there are certain postulates worth remembering: i) There is no relationship between the size of the target population and the size of the sample. ii) The major considerations in determining the size of the sample are (a variability of the populations; (b) the desired

precision; (c) the degree of confidence and (d) the cost of the survey. iii) The greater variability, precision, and confidence the larger should the sample you should take. To double the precision, you should quadruple the sample size. There is geometric relationship, then. iv) Determining the size of a sample involves a trade off between variability, precision and confidence on the one hand, and the cost* survey on the other. v) In statistical theory of sample of size less than thirty is calk small sample. If size is 30 or more, it is statistically a large sample, irrespective of population size. All the above postulates refer to probability sampling. Cost and rule of thumb are used in determining sampling size in nonprobability sampling. It is fashionable to go for a sample of 100, 200, 300 or so. This is a rule of thumb based on seemingly round number sample size. 3.7.2.1 Simple Random Sample for Estimating the Mean

Sample size varies with the purpose of sampling. For estimating mean of population a procedure is dealt below. Let us now assume that the researchers are conducting a random sample survey to estimate the mean income of a population consisting of N retailers in a city. His sample consists of say, 100 retailers (n=100). Suppose the incomes of retailers in the city have a standard deviation of Rs. 800 per month (=800). The standard error of the mean (ax), is that is the standard deviation of the means of all possible samples of size, a, is given by the formula.

X = /n
In our case,

x = 800/1OO = 800/10 = 80

95% of the sample means would fall within 1.96 standard errors on either side of the population mean, as already stated in an earlier section. Suppose our sample of 100 retailers gives mean income of Rs.5000 per month. (X=5000). We can say now that the average income of all retailers in the city would fall in X 1.96 1.96

x. That is X x/n

x.

That is X 1.96

/n.

In this case, the population mean is the

would fall between Rs. 4843.2 and Rs. 5156.8. 1.96 sampling error for a 95% confidence interval.

Suppose you want to estimate the population mean with a precision of 50 rupees, with a 95% degree of confidence 50 - 1.96 Squaring n = [1.96

x/n or = n - 1.96 x/50. x/50]2 = [1.96 x 800/50]2 = (31.36)2 = 983.45 or 983

Suppose the sampling error is *e'. The formula for the size of the probability sample would be n = (z

/e)2, where Z scores relates to the

degree 'confidence, 0 is the standard deviation of the population, e is the degree of vision of the estimate is required. You can note from the above formula that the size of the population, N, does not appear in it. Hence the sample size is unrelated to the population, but related to the confidence level, the std. deviation and the expected degree of its precision. In the formula, z and o are in the numerator and e is in the denominator. If in the above example, the population of the retailers in the city has lower variability of income, o would be smaller and n would also be; a

smaller sample size would then do. And if the variability is greater, the sample size would be larger. Also, the greater the degree of confidence with which you make the estimate of the population mean, the greater would be the sample size. The value of z is 1.96 for a 95% confidence level; it is 2.58 for a 99% confidence level and 6 for a confidence level of 99.999965%. As the degree of precision is higher, V would be smaller, and so the sample size would be higher. If you double the precision, e gets halved and sample size quadruples. The sample size was 983 for a precision of Rs.50. For a precision of Rs.25, the size of the sample would be 3934. It is 4 times (22) of earlier sample size. Getting down to fundamentals with respect to the determination of the size of a probability sample, if previous studies have indicated the standard] deviation or the co-efficient of variation of the population, this can be used determine the sample size. Otherwise we are caught in a vicious circle: To know the population characteristics we have to take a sample; to determine the size of this sample, we should know the population characteristics. Some researchers have found a way out of this. They use what sis others have used for similar studies in the past. This may differ from the size is optimal accordingly to the statistical theory, but they would at least have psychological satisfaction that their sample size is in line with those of other similar studies. There is yet another way out. If the population is normally distributed, difference between the highest and the lowest value in the population would} about six times its standard deviation. We can use this knowledge to mi rough estimate of the population standard deviation. Suppose the retail income are known to vary from Rs. 2600 to Rs. 7400 per month. One sixth this figure, namely Rs. 800 can be taken as a rough approximation of using this figure we can determine the sample size.

3.7.2.2

Sample Size for Estimating a Proportion

For estimating population proportion a procedure for deciding sample is given. Let the study be as to what proportion of the population is in favour of privatization of the public sector industry in India. Here, we should make an assumption about the population proportion in favour of and against the above proposition. Suppose we assume that the population is equally divided about it. That is, p=0.50. the standard error of population is pq/n. Suppose also that the required degree of precision is 3 percentage points or 0.03 and the confidence level is 95%, Then, the sample size can be determined as below. The standard error of deviation:

p = pq/n = (0.5) (0.5) /n - 0.5/n

The sampling error = 0.03 = 1.96 p = 1.96x(0.5/n) = 0.98/n or n = 0.98/0.03 Squaring n= [0.98/0.03]2 =1,067 The following table gives the random sample sizes for estimating population proportion for degrees of precision 1 to 5% on the assumption that the population is equally divided on the issue under consideration for 95% and 99% confidence levels. TOLERATED ERROR 1% 2% 3% 4% 5% CONFIDENCE 95% 9604 2401 1067 600 384 LIMITS 99% 16587 4147 1843 1037 663

It may be noted from the above table also that the sample has to be larger I your estimate has to be precise; and the sample would be larger also if the confidence level the estimate is higher. 3.7.3 Size Of Other Probability Samples With a stratified sampling, a lower size sample would be needed for any specified level of precision compared to a simple random sample. But, in the case of cluster sampling, a larger sample size would be needed for a specified level of error than for a simple random sample. James A. Davis writes in his book, General Social Survey (1973) Codebook, Chicago: National Opinion Research Centre that a national level cluster sample of 1500 cases is roughly-equivalent in terms of precision to a simple random sample of 1000 cases. 3.7.4 AD HOC Methods of Determining Sampling Size To determine the sample size precisely is thus not an easy task. Sampling theory provides some guidelines in this regard for probability samples. There are, however, many ad hoc methods adopted by social researchers. We have already seen one of them, namely, to find similar studies and use their sample sizes as a guide. But the studies should be comparable in terms of the number of groups into which the current sample is divided for purpose of comparison. They should also have achieved a satisfactory level of reliability. Seymour Sudman has given the following able based on over several hundred studies in the United States as a guide to social researchers in his book, Applied Sampling, New York Academic Press, 1976.

TYPICAL SAMPLE SIZES FOR STUDIES OF HUMAN AND INSTITUTIONAL POPULATIONS People or Households Number of subgroup analysis National Regional or Special 500 200-500 1000+ Institutions National Regional or Special 50-200 200-500 500+

None or few Medium 1000 - 1500 200 many 2500 + l000+

1500 - 2000 500 - 1000 500-1000

The following points are worth noting here. For national studies the sample size tends to be larger than for regional studies possibly because of possible variability at the national level than at regional level. Samples involving institutions tend to be smaller than those involving people of households possibly because the former are fewer in number than the latter. Another point to be remembered in this regard is that if the topic of research or the population being sampled is of great interest and if had not been studied before, it is quite possible that even a study based on a small sample (and consequently of low precision) will be of much interest to many. On the other hand if a topic has been studied by many persons earlier, no one would be' interested in our study on the same topic unless this is based on a large national* , sample. Sudman suggests in his book, Applied Sampling, another rule for thumb for a sample divided into groups. Each major group should have a minimum sample of 100 and each subgroup may have a sample size of 20 to 50. To maintain this minimum for each subgroup it may be necessary to adopt disproportionate sampling instead of proportionate.

Sometimes the most important constraint may be money and time. For sexample, if only Rs. 5000 is available for a study and it cost Rs. 20 to interview I a respondent and Rs. 2000 would be required for data analysis, this maximum j sample size would be (5000 - 2000) / 20 = 150 only. The choice then may be /between conducting the study with a less than optimum sample size or taking up |some other study which costs less relatively, 3.8 BIASED SAMPLE A biased sample is one where the sampling error cannot be determined. A sample is biased in the following situations: i) The size of the sample is arbitrarily decided, i.e., neither rule of thumb nor statistical theory is followed. ii) The stratification is not scientifically carried out. Some strata may over lap. That is some exclusive strata are combined or aggregated. Some strata are ironically sub-divided. These strata are thus disaggregated. Aggregating the disaggregateable and disaggregating the aggregateable lead to biased sample. iii) The representation of each stratum in the sample if based on some quota rule, disregarding the size of each stratum, then the sample gets biased. iv) The choice of unit is non-random. Accidental meeting, convenience of getting response, judgmental inclusion or exclusion of a sample unit on some extraneous considerations etc. lead to biased sample.

3.9 SAMPLING AND NON-SAMPLING ERRORS Data collection is beset with errors. These are sampling and non-sampling. The following chart gives a picture of sampling and non-sampling errors.

These errors are explained below. 3.9.1 Sampling Errors The errors which arise due to the use of sampling surveys are know the sampling errors. Even when a sample is random one, it may not be exactly representative of the population from which it is chosen. This is because samples are seldom, if ever, perfect miniature of the populations. However these errors can be controlled. Sampling errors are of two types - biased and unbiased. 3.9.1.1 Biased errors

Biased errors are those which arise as a result of any bias eg. prejudice of the person in selecting a particular sampling method. For example, purposive sampling method may be adopted in place of a simple random sampling method. As a result of such a selection, some errors are bound to arise, and they are known as biased sampling errors, or cumulative errors or non-compensating errors. As bias or prejudice forms a constant component of error that does not decrease

as the number in the sample increases. Instead such errors are likely to increase with an increase in the size of the same. Bias may arise due to. (i) faulty process of selection, (ii) faulty work during the collection of information, and (iii) faulty method of analysis. Faulty selection of the sample may give rise to bias in a number of ways, such as: (a) purposive sampling, (b) selection of sample in a haphazard way, (c) substitution of the selected item in the sample by another, (d) incomplete investigation or response etc. During the process of collecting the actual information in a survey (whether sample or census), certain inaccuracies may creep in. These may arise due to the improper formulation of the problem, wrongly defining the population, specifying wrong decision, securing an inadequate frame, poorly designed questionnaire, an ill-trained interviewer, failure of a respondents memory, unorganised collection procedure, faulty editing or coding of response and so on. Faulty methods of analysis may also introduce bias. Such bias can ho avoided by adopting the proper methods of analysis. Biased sampling errors can be measured and controlled to some extent. 3.9.1.2 Unbiased errors

Unbiased errors arise due to chance factors, between the members of the population included in the sample and those not included. It is known as random error. The random sampling error decreases on an average as the le increases. Such error is, therefore, also known as non-cumulative or compensating error. Unbiased sampling errors can be statistically measured id controlled. 3.9.2 Non-Sampling Errors Non-sampling errors are errors not connected to sampling. This type of error can occur in any survey, whether it be a complete

enumeration or sampling. Non-sampling errors include biases and mistakes. Biases are deliberate while mistakes are not random. Mistakes result from inefficiency. Some of the factors responsible for the non-sampling errors are enumerated here. Vague definition of population, vague questionnaire, vague conception regarding the information desired, inappropriate methods of statistical unit, or inaccurate/inappropriate interview, observation

measurement, errors in data processing operations such as coding, punching, verification, tabulation etc. and errors committed during presentation and printing of tabulated results. QUESTIONS 1. Explain the meaning of sampling design and the basic elements involved in this context. 2. Distinguish i) sampling from sample frame and ii) population from universe and iii) sample survey from census survey. 3. 4. 5. 6. Examine the pros and cons of sampling method. Discuss the various methods of probability sampling. Discuss the various methods of non-probability sampling. Do you consider probability sampling superior to the nonprobability sampling? Why? 7. 8. On some cases non-probability sampling is adopted. Why? What is sample size? Explain the statistical methods available in it estimation. 9. Explain non-statistical bases of deciding sample size.

10.

What are sampling and non-sampling errors? Explain how these can controlled.

REFERENCES 1. 2. 3. 4. 5. 6. 7. Research Materials - P.Saravanavel Elements of Educational Research - S.P.Sukhia Research for Marketing Decisions - Green & Tull Marketing Research - Sangeetha Agarwal Research Methdology - Basetia and Sharma Encyclopedia of Modern Education Research - S.K. Sharma Research Methodology - C.R. Kothari

* * *

UNIT IV

DATA COLLECTION : SOURCES, METHODS AND TOOLS

In this unit Sources of Data, primary and secondary data, Modes of collection, Analytical method, case study, observation, survey method. Interview method and its importance, Types of interview, Preparation for an interview, Effective interview techniques, Limitations of Interview, Schedule and its meaning and kinds, Essentials of a good schedule, Procedure for the formulation of a schedule, Questionnaire, Meaning and types, Format of a good questionnaire, Factors affecting the response to a questionnaire, Advantages and limitations of schedules and questionnaires, pre-testing and its importance, scaling methods, scale construction techniques, etc. are dealt.. Introduction Data collection is the most important of all steps in research. Scientific data collection methods, tested tools of data collection and reliable sources of data are essential for good research outcome. 4.1 DATA FOR RESEARCH Any research work needs data. Data throw light on the problem, its dimensions, the solutions and their effectiveness. Actually data that throw on the problem, solution, etc have to be collected adopting appropriate method, tool and strategy. Take any marketing research arena: product launching, product success, product failure, product line addition, product line deletion, bra equity, price strategy, effect of price-rise on market share, effect of price-off on demand pattern, inventory policy, on warehousing brand equity, policy, logistics of decisions, of advertisement effectiveness advertisement

campaign, behaviour of salesperson in promoting product-range etc.

are certain rest problems. We have to collect data relevant to these arena so that the problem understood well and appropriate solution arrived at. Such solutions aid decision making. All business decisions involve recognizing alternative courses of action to solve problems and making predictions about the behavior of the stakeholders when alternative solutions are tried. Choose any business problem decisions made to solve it will ultimately turn, in whole or in p prediction of the behavior of the participants, who may be: employees, stakeholders, suppliers, consumers, industrial users, mi competitors and the like. Sufficient data base is needed thus to make the predictions. Through research relevant data are collected and relevant data base built. The data needed may be primary or secondary or both. 4.1.1 Types of Data - Primary and Secondary Data Two types of data are used in research. Primary data and second are the types involved. The term data is in plural form; the singular for term is datum. Readers better avoid using the grammatically wrong data wrongly taken as the plural for the term, 'data'. Primary data refer to data that are collected afresh using a specifically designed tool to meet the specific requirements of the current research work or investigation at hand. On the other hand, secondary data are data that are; from works already done by somebody and used for the purpose of the rest The following illustration may serve to clarify the distinction between primary and secondary data. Consider the case of a Company, which wants to evaluate the effect of an advertisement promotion drive for their brand of detergent. In order to estimate the

effect of this promotion, say in New

DI

Madras, the marketing manager

may want to monitor the retail sales to consumers in that area, before and after the promotion is introduced. One way of monitoring sales is to choose a sample of retail stores, and audit their records of sales to consumers every month for the entire period of study. The data collected are primary data because the data are collected exclusively for purpose of the present research work. An alternative method of acquiring required sales data is to obtain the same from the commercial research age like Operations which regularly collects
(

Research

Group

(ORG),

information by means of store audits. If

the Company obtained the data ^ ORG (rather than collect them directly from the retail stores), it would be using secondary data. 4.1.2 Choice between Primary and Secondary Data The investigator must decide at the outset whether be will use primary data or secondary data in an investigation. The choice between the two depends on: (a) nature and scope of the enquiry or study, (b) availability of financial resources, (c) availability of time, (d) degree of accuracy desired, and (e) the status of the investigator i.e., individual, corporation, Government, etc. Primary data generally used in those cases where the secondary data do not provide an adequate basis for analysis. In certain cases both primary as well as secondary data may be employed. The reason why secondary data are being increasingly used is that published statistics are now available covering diverse fields so that an investigator finds required data readily available to him' in many cases. For certain studies like stock price behaviour, interest and exchange rate scenario, etc. only secondary data are used.

4.2

OBTAINING PRIMARY DATA Primary data are obtained' through analytical method, case

study, survey method, observation and interview. These are dealt with here. 4.2.1 Analytical Method One of the methods of data collection is analytical method. To analysed means to dissect to see penetratingly to look into deeply and then synthesis dissected parts, in appropriate forms. analytical method experimentation. i) Single Variable Analytical Method : If two situations are similar every respect and one element is added to or subtracted from one, bi not the other, any difference that develops is the result of the operatic that element added or deleted. This is the law of single variable according to J.S.Mill. Effect of new advertisement on brand image sales in one area only, keeping everything same in both the identical areas is an example. Analytical experiments can be constructed even when more two situations are involved, but only one variable is assumed to be impacting on the phenomena. Such experiment is called completely randomized design. The impact of the single factor cause differences if any, can be isolated by introducing varied levels of the single factors operating on the chosen set of elements. Paired sample studies, oneway ANOVA studies etc. come in the category. ii) Double Variable Analytical Method : When the impact of two factors on the phenomena studied is to be separately studied, completely randomized block design experiment is conducted. Here, varied levels of the two factors are introduced. Suppose The basis of

rows and columns are used to represent the two factors. Then each row reflects a particular level of one factor and each column reflects a particular level of the other factor. Two-way ANOVA experiment is used here. iii) Three variable analytical method : Say, variety of fertilizer, variety of paddy and level of water are studied for their influence on yield Of paddy. A Latin square analysis can be executed here. This is a three variable analytical method of obtaining data and hypothesis testing. 4.2.2 Case Method Case method is one of the methods of investigation. Originally this method was used in medicine. The physician used to maintain a record of treatment and health conditions of each patient. Each such record pertains to a 'case'. The peculiarities of each 'case' are known as well as generalities of all .cases together. The cause-effect relationship is established and exceptions there to are also noted down. According to H.Odum, case study method is a technique by which individual factor whether it be an institution or a group or an episode in life of an individual is analysed. The basic unit of case study is a case - a particular one of its kind. It may be a single business unit or a whole group of business units in an area or a single producer or a group of producers or a single consumer or a group of consumers and so on. When a group is involved, that group is treated as one unit, with whatever phenomenon found in one group member is taken as a phenomenon of the group as such. Characteristics : The important characteristics of the case study method are as ''Under:

(i)

Under this method the researcher can take one single social unit or such units for the study purpose; he may even take a situation to study the same comprehensively.

(ii)

Here the selected unit is studied intensively i.e., it is studied in minute details. Generally, the study extends over a long period of time to ascertain the natural history of the unit so as to obtain enough information for drawing correct inferences.

(iii)

In the context of this method we make complete study of the social unit covering all facets. Through this method we try to understand the complex of factors that are operative within a social unit as an integrated totality.

(iv)

Under this method the approach happens to be both qualitative and quantitative. Mere quantitative information is not collected. Every possible effort is made to collect information concerning all aspects of life. As such, case study depends on perceptions and gives us a clear; insight into life. For instance, under this method we not only study how many new product launches a firm has done but shall peep into the] factors that forced it to launch them when we are making a case study of a firm as an innovator. The objective of the study may be to suggest ways to make firms innovative.

(v)

In respect of the case study method an effort is made to know the mutual inter-relationship of causal factors.

(vi)

Under case study method the behaviour pattern of the concerning unit, studied directly and not by an indirect and abstract approach.

(vii)

Case study method results in fruitful hypothesis along with the which may be helpful in testing them, and thus it enables the generalized knowledge to get richer and richer. In its absence, generalized social science may get handicapped. Steps involved in case study are:

i)

Determine the present status of the phenomenon under investigations through direct observation or measurement. For example, to make a case study of an outstanding business unit, first the investigator has to survey the present status of the business unit.

ii)

Determine the most probable antecedents of the case and formula definite hypotheses or set of hypotheses. Outstanding business success emanates from commitment to quality and customer focus of great order.

iii)

Direct information from documents, diaries of leaders of the unit, records, dreams of leaders, etc. Also direct interview and observation will yield good amount of data.

iv) v) vi)

Verify the hypothesis. Finally follow up the case again to reiterate earlier findings. Completeness of data, validity of data, continuity,

confidential recording and scientific synthesis are the characteristics of a good study.

4.2.3 Survey Method Surveying means contacting for getting certain information. Survey method is a method of collecting data. There are personal surveys, mail surveys, telephone surveys and net surveys. Survey may be sample survey or census survey. The features of each of these are given below: 4.2.3.1 Personal Survey

Personal survey involves meeting personally every number who has to be surveyed. The features of this method of data collection are as follows. i) The number of respondents that can be contacted is not very high, as the time taken to contact the respondent, and the time spent on the interview itself is very high in relative terms. ii) When the time available for research in large the personal method is used. iii) The cost involved is highest in the personal method since it requires field interviewers as well as their conveyance/ traveling costs. Also, if a person is not available he may have to be contacted again and again. iv) The accuracy obtained in very high, as the right persons are contacted and if there is difficulty in their understanding certain questions the interviewer can take care of it Also, if the interviewer feels that the respondent is not furnishing the correct facts, by observing he can make his own interpretations, record the responses for better results. The

responses rate is high compared to the mail survey, .making the accuracy of the results better. v) When a large geographical area is to be covered and the time and cost constraints are high, personal survey method is not resorted to. However, if it is an ongoing syndicated research or census surveys, such in time and costs have to be necessarily borne for the sake of better coverage and accuracy. vi) This method would require the agency to have a good infrastructure of data collection, in terms of field force, its supervision and control. vii) When the literacy levels are low and the respondent would find it difficult to fill up the questionnaire on his own, this method is the best alternative available. viii) A very length questionnaire under a structured survey is difficult to administer personally, with inaccuracy creeping in on account of the monotonous nature and fatigue effect on the part of the interviewer. At times, in certain socioeconomic studies when this is unavoidable, the number of interviewers is made larger and certain incentives may be given to the respondents to extend their co-operation in filling up the set of questions, asked. ix) The availability of skilled interviewers can reduce the interviewer bias on account of recording incorrect responses of the fatigue effects. x) When the questions require spontaneous answers, this is the best method, However, if the questions are of a personal

nature or require too much thought on the part of the respondent he may feel embarrassed or make up the answers without thinking. However, in case of non-structured and non-disguised techniques like the in-depth interviews such probing is called for. xi) The interviewer may have the tendency to contact some other person; similar to the respondent to complete his quota of respondents. This affects the accuracy of results thus necessitating a tight control on field work. xii) A complete list of the respondents would be required to draw a representative sample. However, the interviewer has at times to use his own availability. 4.2.3.2 Mail Survey Method discretion and access respondents with similar demography characteristic in case of non-

Mail survey involves contacting the respondents through post. A questionnaire is dispatched to elicit views. The features of this method are as follows: i) A large number of respondents can be contacted using the available database of addresses. ii) When the time available is again fairly large, and

respondents are very distantly located this method is preferred. iii) The cost involved is not very high and it is mainly just the cost of mailing which is often very cheap.

iv)

The accuracy obtained is not very high since the response rate is not very high, not more than 20%. Also, the right respondent may not have filled up the questionnaire. Again, the chances of interpreting the questions incorrectly by the respondents are high, resulting in wrong responses which may be inconsistent with the responses in the rest of the questionnaire.

v)

The mail method most suited to contact people scattered over large geographical areas when the time available for contact is fairly enough An updated database of respondent addresses would go a long way in obtaining a good response rate. With the availability of fax, internet and courier facilities, time can be saved.

vi)

The size of filed force required is almost negligible and the agency could function with few data entry operators to take care of the data collected.

vii)

The mail method can be used to contact only people who can read and write and definitely has limitations for social research at the slum level or rural levels.

viii)

This is the method best suited when the questionnaire is very length. The respondents does not have to fill it up in one stroke. He also has time to think about the questions and answer them. Errors on account of fatigue or monotonous nature of the dialogue with the interviewer are not likely to creep in.

ix)

The interviewer bias is not of any consequence in this case.

x)

The questions which require spontaneous answers would not lend themselves suitable to the mail survey. However, personal questions or those involving certain through processes are best suited to the mail questionnaire method.

xi)

The questionnaires are out of the control of the agency. The person who fills up the questionnaire may not be desired respondents but someone in the same office or household.

xii)

The mailing would not solve the purpose if the addresses of the respondents have not been updated. The existence of a mailing list is essential.

4.2.3.3

Telephone Survey Method

In telephone survey voice contact is directly established with the respondents. i) The number of respondents who can be contacted is fairly large, as the time to contact them is less than that for a personal interview. ii) When very little time is available for the research the telephone method is found to be useful. iii) The cost involved in moderately high as skilled telephone operators need to be employed. Also if the respondent is not available he needs to be contacted more than once. iv) The accuracy obtained is fairly high in this case as the response rate is comparable to the personally administered questions. In certain cases, it might be higher also. The skill of the telephone interviewer makes the respondent at ease

and comfortable to answer and the questions. Also, certain questions, which need explanations can also be posed through the telephone. v) The telephone method has a major limitation when

outstation contacts are to be made since the costs would be high for such contacts. Also, it can be used for respondents who necessarily own a telephone. But of late, long distance calls are cheaper vi) The infrastructure required in terms of telephones, database of telephone numbers and skilled telephone operators is extensive. vii) The telephone method can be used for respondents having the ability to communicate and express themselves. As such, only a certain class of respondents lend themselves useful to such methods. viii) The length of the questionnaire has to be extremely short in this case. The issue which is addressed through the questions has to be focused one so that, less time is required to be spent over the telephone to explain the purpose of the research etc. The telephone method would thus have a limited use. ix) The skill of the operator/interviewer largely is responsible for the size of bias in recording the respondents incorrectly. x) The questions requiring spontaneous answers or of a personal nature can be administered over the telephone but those which require thinking would be difficult to take care of using this method.

xi)

The person answering the telephone may not be the desirable respondents who might be difficult to contact,

4.2.3.4

Net survey

This is the world of connectivity through net. Net survey involves using internet. The superiority of this survey is that it has no limitations of geography, The questionnaire may be put through a web-site. The web-site may be advertised again through net or other mode inviting the attention of the prospective respondents. They may send their responses again over the net. This is the most modern one and has great potentials. It is not prohibitively expensive. 4.2.4 Observation method The Concise Oxford Dictionary defines observation as 'accurate watching and noting of phenomenon as they occur in nature with regard to cause and effect or mutual relations. Simply, it means the careful and systematic watching of facts as they occur in course of nature. In the strict sense, observation implies more use of the eyes and the ears than the mouth. 4.2.4.1 Aspects of Observation

Observation involves three processes, i.e., (i) sensation (ii) attention (iii) perception. Sensation is gained through the sense organs which depend upon the physical alertness of the observer. The sense organs are receptive to stimuli and get attracted leading to the first stage in observation. Then comes attention or concentration which is largely a matter of commitment and will-power. Adequate training and experience can make it almost a matter of habit. The third is perception which comprises the interpretation of sensory

reports. Thus, sensation merely reports the facts as observed but perception enables the mind to recognize the facts. Through this process, observation serves the purpose of (a) studying collective behaviour and complex social situations; (b) following up of individual units composing the situations; (c) understanding the whole and the parts in their interrelation; (d) getting the out of the way details of the situation. . 4.2.4.2 Types of Observation

There are different types of observation. Prof, P. Saravanavel has given a good list of the same. i. Casual and Scientific Observation An observation may be either casual or scientific. Casual observation occurs without any previous preparation. It is a matter of chance that the right thing is observed at the right time and in the right place. Scientific observation, on the other hand, is carried out with due preparations and is done with the help of right tools of measurement experienced enumerators and under able guidance. Scientific observations yield thorough and accurate data. ii. Simple and Systematic Observation An observation may be either Simple or Systematic. Simple Observation. is found in almost all research studies, at least in the initial stages of exploration. Its practice is not very standardized. It befits the heuristic nature of exploratory research. Participant studies are also usually classified as simple observation because participant roles do not permit systematic observation. Systematic observation, on the other hand, employs standardised procedures, training of observers, schedules for recording and other devices to control the

observer sometimes even the subject. Clearly some systematisation is valuable in research observation, but the situation often limits what can be done. A systematic observation is a scientific observation too. iii. Subjective and Objective Observations An observation may be either Subjective or Objective. In every act of observation there are two components namely, the object (or what is observed) and the subject (or the observer). It may be that some times one may have to observe one's own immediate experience. That is called Subjective Observation or Self-Observation or introspection. Prejudices and biases are generally parts of subjective observation. Many data of psychological interest are gathered by the method of subjective observation. To avoid such prejudices, the observer takes stock of him and discovers what prejudices and biases will prevent impartial study and disinterested points of view. Persistent self-observation and criticism by others may ultimately overcome prejudice and biases. Such introspection may have another social value i.e., it sensitises the observer to the problems of others and creates sympathetic insight which facilitates, at least to some degree, the understanding of people's behaviour in similar circumstances and similar cultural contexts. The net result is impartial subjective observation. When the observer is an entity apart from the thing observed, the observation of this type is objective . iv. Factual and Inferential Observation Observation may be either factual or inferential In factual observation things or phenomena observed with naked eyes are reported. In inferential observation behaviour or psychological aspects are observed.

v.

Direct and Indirect Observation Observation may be either Direct or Indirect. In the case of

direct Observation the observer is physically present and personally monitors what takes place. This approach is very flexible of events and behaviour as they occur. He is also free to shift places, change the focus of the observation, on concentrate unexpected events if they should occur. In indirect observation recording is done by mechanical, photographic or electronic means. For example a special motion picture camera which takes one frame every second is mounted in a department of a large store to study customer and employee movement. vi. Behavioural and Non-behavioural Observations Observation may be either behavioural or non-behavioural. As pointed earlier the concept of observation involves not only watching but also listening and reading. Thus, observation includes the full range of monitoring behavioural and non-behavioural activities and conditions. Non-verbal analysis, linguistic analysis, extra-linguistic analysis and spatial analysis are the four major categories of behavioural observational study of persons. Record analysis, physical condition analysis and physical process analysis are the three major categories of non-behavioural study of persons. Non-verbal behavioural observation includes observation of body movement, motor expressions and even exchanged glances. Body movement, is an indicator of interest or boredom, anger or pleasure in a certain environment. Motor expressions such as facial movements can be observed as a sign of emotional studies. For instance, eyeblink rates are studied as indicators of interest in advertising messages. Finally, exchanged glances might be of interest in studies of interpersonal behaviour. Linguistic behaviour is a second frequently used from of behavioural observation. One simple type, familiar to most students,

is the tally of 'ahs' (or other annoying words or sounds) that a professor emits during a class. 4.2.5 Interview Method Interview is one of the popular methods of data collection. The term interview can be dissected into two terms as, 'inter' and 'view.'. The essence of interview is that one mind tries to read the other. The interviewer tries to assess the interviewed in terms of the aspects studied or issues analysed. 4.2.5.1 Purpose and Importance of Interview

The main purposes of interview as a tool of data collection - field procedure - is to gather data extensively and intensively. As Pauline V.Young pointed out that the objectives of the interview may be exchange of ideas and experiences, eliciting of information pertaining to a very wide range of data in which the interviewee may wish to rehearse his past, define his present and canvass his future possibilities. Thus, in brief, the objectives of interviewee are two fold: (i) to exchange ideas and experience and (ii) to elicit information. The importance of interview may be known through these points, i) ii) iii) iv) It is the method best suited for the assessment of personal qualities. It has definite values for diagnosis of emotional problems and for therapeutic treatments. It is one of the major bases upon which counseling procedures are carried out. It provides information to supplement other methods of collecting data.

v)

It may be used, in addition to observation, to verify information obtained through correspondence methods.

4.2.5.2

Types of Interview

There are different types of interview. An interview is either structured or unstructured, depending upon whether a formal questionnaire has bean formulated and the questions asked in a prearranged order or not. An interview is also either direct or indirect as a result of whether the purposes of the questions asked are plainly stated or intentionally disguised. Cross-classifying these two characteristics provides four different types of interviews. That is, an interview may be: (1) structured and direct, (2) unstructured and direct, (3) structured and indirect, or (4) unstructured and indirect. Types (1) and (2) are basically objectivist types; (3) and (4) are subjectivist types. There are other types of interviews, like focus-group interview, depth interview, etc. All these are dealt here. i Structured-Direct Interview The usual type of interview conducted during a consumer survey to obtain descriptive information is one using a formal questionnaire consisting of non-disguised questions, a questionnaire designed to "get the facts". If the marketing search manager of a television set manufacturer wants to find out how many and what kinds of people prefer various styles of television cabinets, for example, he may have a set of questions drawn up that asks for these facts directly. Assuming that personal interviewing is being used, each interviewer will be instructed to ask the questions in the order given on the questionnaire and to ask only those questions. The resulting interviews will be structured-direct in nature.

ii.

Unstructured-Direct Interview In the unstructured-direct method of interviewing, the

interviewer is given only general instructions on the type of information desired. He is left to ask the necessary direct questions to obtain this information, using the warding and the order that seems most appropriate in the context of each interview. Unstructured-direct interviews are often used in exploratory studies. Many research projects that use a formal questionnaire for the final interviews go through an exploratory phase in which respondents are contacted and unstructured interviews are held. These interviews are useful in obtaining a clearer understanding of the problem and determining what areas should be investigated. iii. Structured-indirect interview In the case of structured indirect interview the questions are pre-decided and arranged in a structured way. However the purpose of the study is not revealed. iv. Unstructured-indirect interview In the case of unstructured indirect interview the questions aren't pre-decided and neither the purpose of the study made known explicitly. v. Focus-Group Interviews Perhaps the best-known and most widely used type of indirect interview is the one conducted with a focus group. A focus-group interview is one in which a group of people jointly participate in an unstructured-indirect interview. The group, usually consisting of 8 to 12 people, is generally selected purposively to include persons who

have a common background or similar buying or use experience that relates to the problem to be researched. The interviewer, moderator, as he or she is more often called, attempts to focus the discussion on the problem areas in a relaxed, nondirected manner. The objective is to foster involvement and interaction among the group members during the interview will lead to spontaneous discussion and the disclosure of attitudes, opinions, information on present or prospective buying and use behavior. vi. The Third-Person Technique The simplest way of obtaining information through indirect questioning of a respondent is to ask for the view of a neighbor, an (unnamed) associate, or some other person whose views on the subject at hand might reasonably be known. This permits the respondent to project his own views with no feeling of social pressure to give an "acceptable" answer. vii. The Depth Interview There is substantial use of the unstructured, informal interview in marketing research to explore the underlying predispositions, needs, desires, feelings, and emotions of the consumer toward products and services. This method of interviewing is referred to as a "depth interview". The depth interview in marketing research may consist of either direct or indirect questions, or some combination of the two. The skilled interviewer will generally employ both types of questions, A direct, free answer question such as "What are the major reasons why you bought your CD radio? Might well be followed up, for example, with an indirect question such as "Why do you think people who own CD radios bought them?" By following leads and cues provided by respondents, phrasing questions to continue the flow and pattern of the conversation and to maintain the rapport established,

the competent interviewer can explore and probe the underlying motivations of the respondent. viii. The Personal Interview As the name implies, the personal interview consists of an interviewer asking questions of one or more respondents in a face-toface situation. The interviewer's role is to get in touch with the respondent(s), ask the desired questions, and to record the answers obtained. The recording of the information obtained may be done either during or after the interview. In either case, it is a part of the interviewer's responsibility to ensure that the content of the answers is clear and unambiguous and that it has been recorded correctly. ix. The Telephone Interview Telephone interviews are sometimes used in lieu of personal interviews, especially when the information must be collected quickly and inexpensively and the amount of information required is limited. The telephone interview is well suited to such research problems as determining "coincidental" viewing of television or listening to radio programmes. In this type of study, calls are placed ^to a sample of telephone subscribers during the time the programme is on the air. The person received the call is simply asked "Are you now watching television?" and, if so, "What programme you are watching?" Other questions such as "How often do you watch this programme?" "Who sponsors this programme?" and the like may also be asked. The result is a rapid and inexpensive measurement of audience level. Either a structured or an unstructured interview may be held. Since the amount of information sought is usually well defined, non-confidential in nature, and limited in amount, virtually all telephone interviews are structured in nature. This medium does not lend itself well to indirect interviews and has not been used for this purpose.

4.2.5.3

Effective interview Techniques

There are three major processes or techniques of interviewing in which the research worker should develop proficiencies: (a) creating a friendly atmosphere, (b) asking appropriate questions, and (c) obtaining right, responses. These are discussed in detail by Bingham and Moore, Garrett and Fenalson. However, some of the basic rules of interviewing are follows: In establishing friendly relations, spend some time at it to develop a cordial setting for the interview. Engage in pleasant conversation through which questions can arise naturally. Try to establish a relationship of mutual confidence and frankness by stating the purpose of the interview clearly and simply with sincerity and reasonableness, linking the topic of the inquiry to interest of the person. Establish pleasant associations or circumstances. Help the interviewee to feel at case and ready to talk. Be at ease yourself and make it apparent. Allow him time to get acquainted with the surroundings and to gain poise. Make him feel that he is an equal who can exchange ideas and opinions with you. Avoid letting the interview take on the character of an inquisition or "third degree". Avoid, also making it an oral questionnaire. In getting to the essential points of the interview; begin with pleasant topics and avoid unpleasant associations and emotional attitudes. Ask questions, at first that are not likely to cause refusal to answer or to provoke any form of 'negativism. Begin with topics that are neutral with respect to emotional reactions land gradually lead into more personal and emotionally colored questions as the interview progresses. Be straightforward and frank rather than shrewd or clever. The interviewee is likely to detect any cleverness or deviousness on the part of the interviewer and answer in like manner.

Approach the problem as soon as rapport is assured, but do not ask pertinent questions until you think he is ready to give the information accurately and willingly. To facilitate responses, practice taking the interviewee's point of view -think in his place. Ask only one question at a time and avoid implying the answer to your own question. Focus his attention on the question and make sure that he understands what is wanted. Listen, give him a chance to talk, let him tell his story, and then help him to supplement it. Observe his behaviour, facial expressions, tones of voice, and other indirect responses. Allow time enough for him to answer one question fully, but prevent him from rambling, by keeping control of the interview without being domineering. Keep on the subject but listen carefully for casual remarks that might be revealing. If unusual facts are revealed, do not betray surprise, shock, or emotional tensions at the disclosures. Face the facts professionally and avoid excessive sentimentality, sympathy, or antipathy. In the entire process, avoid taking the role of "teacher", avoid a patronizing attitude, do not embarrass the interviewee unnecessarily, avoid putting him on the defensive or arousing feelings of antagonism, and keep from making the interview merely a social visit. In phrasing questions, be careful not to offer alternative questions that might imply what you think is the correct answer - do not "put words in his month". If the interview becomes time consuming, arrangements should be made for another interview at a later time rather than extending the present session and obtaining superficial responses or creating a negative attitude. 4.2.5.4 Pre paring for interview

The following guides should be followed in planning and carrying out interviews.

i)

Determine who is to be interviewed It is necessary in the selection of interviewees to determine who

has information necessary for the research study. Are they in a position to divulge the information and are they willing to give the information that is asked for'.' These individuals must be chosen with care. The researcher should attempt-to .find out as much as possible about the interviewee's interests, his connections, acquaintances, what he has written in the area of the investigation, his experience in the fields, etc. It also important to determine at this time the number of interviewees necessary to get reliable information and whether or not they are representative of all the factions or subdivisions of the entire group that is being studied. ii) Make Preliminary Arrangements for the Interview A definite appointment should be made for the time and place of the interview. The research should seek a time that is satisfactory to the interviewee and meet the appointment promptly. He should arrange for a place of meeting that provides privacy and an atmosphere conducive to soliciting the interviewee's responses. It is also desirable, at times, to facilitate contacts through intermediaries, such as mutual friends or acquaintances, letters of recommendation or authority, or personal letters to create an interest in the problem being studied. iii) Determine the plan of the interview and the questions to

be asked. In preparing for the interview, it is necessary to decide just what is expected to be accomplished, the objectives of the interview, what facts to bring out, what information is to be solicited, what attitudes

are to be established, and what actions are to be motivated. In doing this, the researcher should know his fi^H.have his problem well formulated, and have a schedule or list of questions prepared. This step in the interview method is practically identical to the second guide for the observation method. iv) Conduct a Preliminary Tryout of the Interview Plan Before carrying out an interview with a subject from whom information is desired for the study, the researcher should try out his plan of conducting the interview with friends, classmates, or other persons not in the interviewing sample but like those from whom the information is desired. This procedure will aid the researcher in improving his techniques of asking questions, in reducing the amount of ambiguity or misunderstanding in the way questions are asked, f and should provide some indication of the kind or responses he likely to obtain. 4.2.5.5 Limitations of Interview The interview method has certain limitations. These are as follows: (i) Uneconomical : The transportation cost and the time required to cover addresses in a large area as also possibility of nonavailability or 'not at home', may make the interview method uneconomical and often unfeasible. (ii) Emotionalism : The interview is liable to cease being objective and becoming emotional. In the current of emotionalism all objectivity is swept aside. (iii) Personal bias : Because of the personal bias the data may be distorted, resulting in wrong generalizations. In spite of the best efforts the personal bias cannot be totally eliminated in this

method. There is likelihood for data contamination through rumuors or circulation of questionnaire items. Interviewers subjective adjustment to village conditions may lead to selective recording of local events. They may tend to rely on a typically accessible informants. This method takes time and social adaptability to develop rapport. Supervision in this method is time-consuming, irregular and expensive. Even if a superior visits each village once a week, he never knows what is happening in the interim. 4.2.5.6 Merits of Interview

Despite the expensiveness and emotionalism of interview, it has certain merits. Intensiveness, thoroughness, situationalism, etc. are involved in interviews. More is learnt not only from words but also from gestures. 4.3 SCHEDULE : MEANING AND KINDS Schedule is the tool or instrument used to collect data from the [respondents while interview is conducted. Schedule contains questions, statements (on which opinions are elicited) and blank spaces/tables for filling up the respondents. The features of schedules are : (i) The schedule is presented by the interviewer. The questions are asked and the answers are noted down by him. (ii) The list of questions is a more formal document, it need not be attractive. (iii) The schedule can be used in a very narrow sphere of social research.

The main purposes of

schedule are three fold : (i) to provide a

standardized tool for observation or interview in order to attain objectivity; (ii) to act as memory tickler i.e., the schedule keeps the memory of the interviewer/ observer refreshed and keeps him reminded of the different aspects that are to be particularly observed, (iii) to facilitate the work of tabulation and analysis. There are several kinds of schedule. Rating schedules,

documents schedules, survey schedules, observation schedules and interview schedules. Rating Schedules is a schedule used to obtain opinions, preferences etc, respondents over statements on the phenomenon studied. The schedule consists of positive and negative statements of opinion on the phenomenon. Documents Schedules are used to collect data/information from recorded evidences and/or case histories. Here the blanks, functional issues related blanks and the like to be filled up from records and documents are present. Survey Schedules are like questionnaires. Observation Schedules are" schedules used when observational method of data collection is used. These could be structured or unstructured Interview schedules are used for collecting data when interview method of communication with the respondents is used. 4.3.1. Essential of a good schedule A good schedule must have the following features (i) Content: Should cover questions or statements relating to all significant aspects of the study. (ii) Dissectional: problem. Should look into the problem analytically,

dissecting every, major and significant components of the

(iii)

Context: Should suit the context in which it is applied. Different types of studies need different schedules.

(iv)

Criterion : Should use sound logic in classifying respondents based opinions expressed.

(v)

Construction : Should be constructed in such a way that questions statements progress gradually and in order. Better it is sub-divided into parts, each part deeding with a certain sub topic of the issue studied. For each objective, a separate part may be devoted.

(vi)

Language: Should be linguistically superbly designed. Clear and straight forward language be used.

vii)

Reliable: Should be reliable such that same results are obtained whenever the schedule is used when everything else remains same.

viii)

Mechanical aspects : Paper used, margin space given, spacing, printing, size of letters, etc. should be normal.

ix)

Size : Should not too length nor too short. Should give fair coverage to the topic.

(x)

Qualities to be avoided: Long, complex, presumptuous, personal, embarrassing, hypothetical issues, morality oriented, upsetting type and necessary questions must be avoided. To sum up, accurate information and accurate response are the

two essential conditions of a good schedule. Accurate communication is effected by proper wording of questions so as to produce desired sense without any ambiguity. Accurate response is said to have been

achieved when replies contain the information sought for. The response is achieved by stimulating the respondents to fill the schedule. Besides, the physical structure of the schedule should be attractive; the questions asked or information sought should be adequate and relevant to the enquiry, so that final generalization may be based upon it The information sought should not only be valid, it should also be capable of being tabulated and if possible being subjected to statistical analysis. 4.3.2. Procedure for formulating a schedule (i) Study the different aspects of the problem. The problem under study should first of all be split up into various aspects. The determination of these aspects will depend upon clear understanding of the problem under study. (ii) Sub-divide the problem to get necessary information. Each aspect has again to be broken up into a number of subparts. These sub-parts should be quite exhaustive to give a full and complete picture of-the aspect under study. (iii) Class questions. Care should be taken to see that the questions convey the exact sense. Respondents will be willing to supply information without any hesitation, bias or distortion of facts, if questions are exact and clear. (iv) Serialization of Questions. In order to obtain well-organised information, it is necessary that the questions should be presented to the respondents in a well-ordered serial. It has been experienced to various field studies that the change in the order of questions affects the answers adversely.

(v)

Testing the validity of schedule Whatever may be the degree of precaution taken, some slips are based to be left out and these can be located when the schedule is put into a reliability and validity test.

(vi)

Division. The schedule be divided into adequate number of divisions. Introductory part, instructional part, issues related parts, etc. are certain parts by which the schedule is divided into parts.

(vii)

Appropriate form of questions. Use appropriate forms of questions at appropriate places. Open ended, close ^ended, pictorial, Yes or No (Questions), multiple choice questions, etc. be used.

4.3.3

Merits of schedules Schedules can be administered even with unlettered people

100% response is there. No misinterpretation results. There is no problem of unintelligible handwriting or choice of words. The situation is learnt and many new chance findings are got. 4.3.4 Demerits of schedules

Not many can be covered. It is expensive. Causes practical problems of logistics and time matching. It is time consuming. Pressures on the respondent to complete and hand over on the spot exist. Since confidentiality is not ensured* frank responses may not forth come. 4.4 QUESTIONNAIRE A questionnaire is a form prepared and distributed to

respondents secure responses to certain questions. It is a device for

securing answers questions by using a form which the respondent fills by himself. It is systematic compilation of questions that are submitted to a sample drawn the population from which information is desired. It is an important instrument in normative survey research, being used to gather information 'from Widely scattered sources. The questionnaire procedure normally comes into use where one cannot readily see personally all of the people from whom he desires responses or where there is no particular reason to see them personally. Purposes of questionnaire are two fold: (i) [To collect information from the respondents who are scattered in a vast area and] (ii) to achieve success in collecting reliable and dependable data. 4.4.1 Types of Questionnaire There are diverse forms of questionnaire. These are discussed briefly here. i. Structured and Non-structured questionnaires The structured questionnaire contains definite, concrete and direct questions, whereas non-structured questionnaire may consist of partially completed questions or statements. A non-structured questionnaire is often used as the interview guide, which is nondirective. The interviewer possesses only a blueprint of the enquiries and he is largely free to arrange the form or statements of the questions. The enquiries framed in a general form beforehand are giving a specific form during the actual process of interview. ii. Closed form and Open Form The questions that call for short or check responses are known as restricted or closed form type. This provide for making a yes or no,

a short response, or checking an item from a list of given responses. It restricts the choice of response for the respondent. He has simply to select a response out of supplied responses and has not to frame his response in his own way. It is easy to fill out, takes less time, keeps the respondent on the subject, is relatively more objective, more acceptable and convenient to the respondent, and is fairly easy to tabulate and analyse. The open form, open-end or unrestricted type questionnaire calls for a free response in the respondent's own words. The respondentframes and supplies his own response. No clues are provided. It probably provides for greater depth of response. The subject reveals his mind, gives his frame of reference arid possibly the reasons for his responses. This type of item is sometimes difficult to interpret, tabulate and summarize in the research report. When the respondent is allowed freedom of response his expressions-may take any unique direction which may not find any uniformity with oilier responses. The mixed questionnaire. The mixed questionnaire consists of both close-end and open-end type questions. For social research, this method is very useful. Many questionnaires include both open and closed type items. Each type has its specific merits and limitations and the research worker has to decide which type is more likely to supply the information he wants. iii. Fact and Opinion Questionnaires Questionnaire are also classified as: (i) Questionnaire of fact, which requires certain information of facts from the respondent without any reference to his opinion or attitude about them, and (2) Questionnaire of opinion and attitude in which the informant's opinion, attitude or preference regarding some phenomena is sought.

iv.

Pictorial and verbal Questionnaires In the pictorial questionnaire, pictures are used to promote

interest in answering questions. It is used extensively in studies of social attitudes and prejudices in children or illiterate persons. In a pictorial questionnaire, the selected alternative answers in the form of pictures are given and the respondent is required to tick the picture concerned. This questionnaire may be very useful for collecting data in a developing country like India, specially from the rural masses who are mostly illiterate and less knowledgeable. The serious limitation of this questionnaire is that it is lengthy in form. Also it is highly expensive. Verbal questionnaire uses words and numbers only. It is the usual form meant for literate respondents. In the questionnaire technique, great reliance is placed on the respondent's verbal report for data on the stimulus experiences in which he is exposed and for knowledge of his behaviour. The questionnaire is effective only when the respondent is able or willing to express his reactions clearly. A good questionnaire can elicit cooperation of the respondent to get frank answers on almost any subject, even such personal matters as sex and income. Thus, it is clear that the respondent can judge the study only by what he can see. The questionnaire, by its very nature, is an impersonal technique and it is several pieces of paper appeals/persuades the respondent that he ought to participate. 4.4.2 Process of Questionnaire Technique The process of questionnaire is divided into (a) Designing or making (b) Issuing (c) Returning. While designing attention should be paid to three important matters; (i) Appearance of the questionnaire, (ii) Contents of questionnaire, and (iii) Subject-matter.

4.4.2.1

Design of questionnaire

To design a good questionnaire certain steps are needed. These are : i) Consult colleagues, friends & acquaintances to get 'their thinking on the problem. (ii) iii) Formulation of such a list of areas & questions. Submit this list to experts both in the field of problem nd related fields. iv) v) vi) Drafting the questionnaire (first stage) Pre-testing the questionnaire. Re-examination and Revising the questionnaire (second stage). Physical Appearance of the Questionnaire

4.4.2.2

Special attention should be paid in order to give a good physical appearance to the questionnaire so as to get good response from the respondent. Matters requiring special attention include quality and size of paper, type of printing/type letters, margin, spacing, arrangement of items on the questionnaire etc, 4.4.2.3 Guidelines on Questionnaire Contents

Any questionnaire must be limited in its length and scope. That is, it is usually unwise to expect returns from a questionnaire which requires much more than 10 to 25 minutes, to complete, and to its use is restricted to rather narrow areas of data. Length itself need not be guiding factor in the design. The questionnaire should cover the

subject, free from unnecessary repetition, and that the techniques used are those which will meet the demands of study. The language should be simple; use words of precise (not flexible) meaning; useful to insert a transitional phrase; at the end, use an expression of appreciation for the efforts put forth by the respondent; use cross tabulations and mock tables. To make the entire schedule or questionnaire a unity (as a unit), there should be a logical progression such that the respondent is (1) drawn (into the interview) by awakening his interest (begin with relevant or harmless questions) (2) easily brought along from items which are simple to answer to those which are complex (3) not affronted by early and sudden request for personal information and (4) never asked to give an answer which could be embarrassing without being given an opportunity to explain. A final question may be used to make a socially appropriate ending to it or final question may be request for an evaluation of the questionnaire or comment of thanks for cooperation at the end of the form. The questionnaire should have the features of spontaneity and flow, comparability, reliability, stability, validity etc. According to Pauline V.Young, the following items are mandatory for most forms of the questionnaire: i) The identification of the agency or organisation collecting the data should appear in a prominent place on the front of the form. If the information is collected by one agency and sponsored by another, the names of both agencies should be shown. If the forms are to .be returned by mail, the address to which they are to be sent must be clearly specified.

ii)

The title of the study or survey should be on the front page of the form and usually it is desirable to have it appear in large print. Printing and spacing devices can both be used for obtaining effective emphasis on the major items of the form.

iii)

The authority for basis or the collection of the information should be stated.

iv)

The confidentialness of the data should be made clear in such a way that the respondent will be assured of protection.

v)

The data of period covered by the report should be stated in advance, or space should be left for them to be filled in by the respondent or the interviewer.

vi)

If footnotes are needed, the space for recording them should be clearly identified and the amount of space should be determined by the expected number and type.

vii)

A place should be provided for the signature of the respondent or the interviewer, unless none is required.

viii)

It is usually advisable to assign a serial number to each questionnaire in order to identify it easily, and to facilitate the control" of interviewing assignments.

ix) s x)

When a questionnaire contains more than one page, each page should be numbered. If the interviewer is permitted to use extra forms, an item should be inserted for the purpose of identifying each from used.

Dilbert C.Miller in his Handbook of Research Design and Social Measurement provides a guide to the questionnaire construction. The guide has the following sixteen 'commandments' to be followed in the formulation of questions: (i) (ii) (iii) (iv) Keep the language pitched to the level of the respondent. Try to pick words that have same meaning for everyone. Avoid long questions. Do not assume that your respondent possesses factual or first hand information. (v) In forming a question, either suggest all possible

alternatives to the respondent or do not suggest any. (vi) (vii) Protect your respondent's ego. If you are after unpleasant orientations, give your respondent a chance to express his positive feeling first so that he is not put in an unfavourable light. (viii) Decide whether you need a direct question, an indirect question, or an indirect followed by a direct question. (ix) (x) (xi) (xii) Decide whether the question should be open or closed. Decide whether general or specific questions are needed. Avoid ambiguous wording. Avoid biased or leading questions.

(xiii)

Phrase questions so that they are not unnecessarily objectionable.

(xiv)

Decide whether a personal or impersonal question will obtain the better response.

(xv)

Questions should be limited to a single idea or a single reference".

(xvi) 4.4.2.4 i)

Establish the frame of reference you have in mind Advantages of the Questionnaire The merits of questionnaires are listed below:

Economical Since the questionnaire approach makes it possible to cover, at

the same time, a large number of people spread over a large territory, it is decidedly more economical in terms of money, time and energy. Other methods do not afford such a facility. ii) Time saving Besides saving money, the mailed questionnaire also saves time. Simultaneously, hundreds of persons are approached through it whereas if they are to be interviewed it may take a very long time. iii) Suitable in Special type of response The information about certain problems can be beat obtained through questionnaire method. For example, the research about sexual habits, marital relations, dreams etc., can easily obtained by keeping the names respondents anonymous. As the respondents are not required to indicate their names on the questionnaires, they feel

free to express of questionnaire method. In view of its anonymity the questionnaire is more effective than an interview. iv) Less pressure on the respondents The method places less pressure on the subject for immediate response, and gives more time to the respondents for properly answering questions. v) Uniformity The questionnaire, by its very nature, is an impersonal technique. Uniformity from one measurement situation to another is provided by virtue of its standardised wordings of questions, standardised sequence of questions and fixed of standardised instructions for recording responses. vi) Vast area coverage The mailed questionnaire is popularly used when the field of research is vast and the respondents are scattered over a very large area. By this method, a large sample may be drawn and all groups of people can easily be covered and contacted. 4.4.2.4 Disadvantages/Limitations of the Questionnaire The following are the demerits of questionnaires i) Illiterate and less educated One of the major limitations of the questionnaire is that it can be administered only on subjects with a considerable amount of education. Complex questionnaire requiring elaborate written replies can be used indeed on a very small percentage of population.

ii)

Proportions of returns is usually low In a mailed questionnaire, the proportion of returns is usually

low; it may sometimes be as low as 10% or even less. iii) Misinterprets/unintelligible reply In a questionnaire, if the respondent misinterprets a question or writes his reply unintelligibly, there is very little that can be done to correct this. In this i approach there is no facility for repeating questions, explaining them or seeking Clarification of a particular response. To conclude, questionnaire designing remains a matter of commonsense and experience and of avoiding known pitfalls. It is not as yet, if indeed it ever can be, a matter of applying theoretical rules. Alternative versions of questions lust be rigorously tested in pre-tests and the pilot study. For the absence of hard and fast rules, test of practicability must play a crucial role in questionnaire construction. 4.4.2.5 Format of Questionnaire

A questionnaire must be of a good format. An introductory part giving a brief on the study and the need for the questionnaire with a request to fill up the questionnaire, an identification part eliciting personal data of the respondent, and main part dealing with questions / statements relevant to each of the objectives and hypotheses set forth are needed. A conclusion part in the end may be added. The main part may certain sub-parts, each sub-part dealing with a particular issue studied perhaps, for each objective a separate part may be provided. Questions or statements relevant to an objective may be grouped under one sub-part under a distinct heading.

A model questionnaire is given below. ALAGAPPA UNIVERSITY DIRECTORATE OF DISTANCE EDUCATION STUDY ON FACTORS INFLUENCING ORGANISATIONAL CLIMATE Guide : Dr. M.SELVAM Candidate : C. PRAKASH

We have undertaken a doctoral research work on organization climate. Kindly do provide your frank opinions. Thanks. The information given will be used for research purpose only under strict confidentiality. PART I - PERSONAL DATA (Please tick in the appropriate box, wherever boxes are provided) 1.1 1.2 1.3 Your name (if you please) Designation Pay per month : : : Upto Rs. 8000 1.4 Age : Less than More than Years 50 Years 1.5 1.6 Sex Educational Qualification Technical Non-Technical 1.7 Experience in years (Total) : Upto 5 Yrs More 10 Yrs 6 to 10 Yrs : : Male Graduate Female Post Graduate Rs. 8001-16000 30-40 40-50 Rs. 16001 & above 30 Years Years

PART II FACTORS AFFECTING ORGANISATIONAL CLIMATE

Given below are some statements relating to the factors influencing organizational climate (O.C). You are requested to read each statement and indicate with a tick () the extent to which you agree or disagree with the different factors as contributing to organizational climate. There is no right to wrong view as such, since perception is purely individual. We trust you will give your free and frank opinion. 1 JOB CHARACTERS by feel an creates in Highly Agree Agree Neutral DisAgree Highly Disagree

1.1 Responsibility making important organization O.C. 1.2 Responsibility to others leads one

positive perception of by to

implying accountability negative perception 1.3 Job pressure implying long hours of work or odd hours of work lead to negative perception 2. PEERS CHARACTERS among and to to and peers underlies mutual understanding that contributes leads positive perception 2.2 Generosity buck -passing 2.1 Co-operation

undue expectation and that lead to negative perception. 2.3 Competition peers is in among the

organizational interest and that contributes to positive perception 3 SUBORDINATES CHARACTERS 3.1 Subordinate obedience brings the out superior better and understanding between subordinate leading to positive perception 3.2 Willingness subordinate excellence 3.3 Non of to the share to of to needs negative

his job knowledge and leads positive perception receptiveness subordinates organizational lead 4. to perception SUPERIOR'S CHARACTERS 4.1 Superior who emphasizes production is a cause for negative perception

4.2 Superior who delegates authority 4.3 A contributes who for to positive perception superior concern attaches

people is a cause for positive perception 4.4 Supportive guidance puts the subordinate in the right trace, leading to positive perception 5 TOP MANAGEMENT with positive with low CHARACTERS 5.1 Leadership leads to simplicity and humility perception 5.2 Leadership concern for production and high concern for people leads to positive perception 5.3 Leadership production perception 5.4 Achievement orientation
of the leader creates a sense of group commitment leading to positive perception

with and for

moderate concern for people leads to positive

6. 6.1

SIZE AND STRUCTURE Large organizations tend to lose elements of personal closeness and team work leading to negative perception

6.2

Bureaucratic structure with rigid hierarchical rules and controls lead to negative perception

6.3

Organisational structures with many layers results in close supervision and lack of employee leading to perception freedom negative of

6.4

Simple

forms

organisation and / or autonomous units with functional responsibility lead to positive perception 7. 7.1 CONFLICT RESOLUTION Organisational conflicts are not conclusive evidences of destruction but signs of development, they do contribute to

positive perception 7.2 A management that is always ready to speak with the employees to solve perception 7.3 A co-operative approach on the part of management in solving conflicts leads to positive perception 8. 8.1 REWARD SYSTEM More than money, reward in the form of enhanced authority and responsibility lead to better perception 8.2 Rewards linked to efficiency and productivity lead to better perception 8.3 Systematic reward system as against adhoc ones contribute to positive perception 8.4 High monetary reward leads to better perception 9. 9.1 OPPORTUNITY FOR GROWTH Opportunity for conflicts contributes to positive

personal growth in the form of promotion and recognition contribute to positive perception 9.2 Career oriented training given by the employer leads to better perception 10. CORPORATE CULTURE 10.1 Homogeneity of values, beliefs and attitudes among members contribute to positive perception 10.2 Less employee turnover resulting in stability of membership and longtime mutual relationship leads to positive perception 10.3 Higher the degree of conformity between the value systems of the employee and that of organisation better the perception

PART III DEGREE OF PREVALENCE OF FACTORS AFFECTING ORGANISATIONAL CLIMATE Given below are parameters of O.C. We request you to state the level of presence of these parameters with reference to your position in your organisation. Here again your perception of presence of these factors count much. Hence what you feel about these is what we need. Kindly feel free to share your view. Very high High 1. a) b) 2. a) b) 3. a) b) 4. a) b) 5. a) b) 6. a) JOB CHARACTERS Responsibility Authority PEERS CHARACTERS Co-operation Generosity SUBORDINATES' CHARACTERS Obedience towards superiors Willingness share SUPERIOR'S CHARACTERS Concern for production Concern for people TOP MANAGEMENT CHARACTERS Concern for production Concern for people SIZE AND STRUCTURE Level of decentralization to job knowledge Medium Low Very Low

b) 7. a) b)

Level of Centralization CONFLICT RESOLUTION Employees opportunity for being heard Presence Management operation of co-

8. a) b) 9. a) b) 10. a) b)

REWARD SYSTEM Job enrichment based by on FOR and authority Rewards efficiency OPPORTUNITY GROWTH Promotion recognition Training for development CORPORATE CULTURE Employee turnover Profit orientation 4.4.2.6. Factors affecting response for a questionnaire Several factors influence the response for a questionnaire. These are: (i) Characters of the respondents: Their interest in the study, their desire to help the researcher, their commitment etc. influence.

(ii)

Name and fame of the guide: The popularity of the guide influences responsible.

(iii)

Prestige of the institution: The prestige of the institution where the research is undertaken influences response rate.

(iv)

Significance of the problem: Problems that touch the life of most of the respondents influences response.

(v)

Features of the questionnaire : A well drafted questionnaire evokes better response. The lucidity of the language used, the sequence of questions, the format of questions, the quality of questions, etc. influence response rate.

(vi)

Length of the questionnaire : Short questionnaires get good response and long ones get less response.

(vii)

Covering letter & introduction : An exceedingly appealing covering letter and introduction about the significance of the study get good response.

(viii)

Guarantee of Anonymity: Guarantee of anonymity increases response rate.

(ix)

Respondent convenience factors ; A stamped, self-addressed envelope to return the filled up questionnaire, complements for the respondents, convenient size of questionnaire papers, follow up reminding the respondents, etc. influence response late.

(x)

Information sought; Type of information sought, amount of information sought, quality of information sought, and the like influence response rate.

(xi)

Time at the disposal of the respondent: This is very important factor.

(xii)

People race against time. Adequate time must be available and be given to the respondent to get more response.

(xiii)

Type of responses needed: Responses that involve mostly tick marks ensure higher response rate.

(xiv)

Literacy level and understanding level of the respondents influence response rate.

(xv)

Personal calls. As a last resort personal calls on the respondents either by the researcher or on his/her behalf will ensure good response.

4.5

TECHNIQUES COLLECTION

OF

COLLECTING

QUALITATIVE

DATA

Certain data are qualitative in nature , Data on attitude, opinion or behaviour of employees, customers, and sales persons etc., are qualitative. The terms "attitude" and "opinion" have frequently been differentiated in psychological and sociological investigations. A commonly drawn distinction has been to view an attitude as a predisposition to act in a certain way and an opinion as a verbalization of the attitude. Thus, a statement by a respondent that he prefers viewing color to black-and-white television programs would be an opinion expressing one aspect of the respondent's attitude toward colour television. Motivation, commitment, satisfaction, leadership effectiveness, etc involve attitude measurement based on revealed opinions .These qualitative data require measurement scales for being measured as no scales are existing. In this section we will

deal with scaling techniques. New scales and method of measurement have to be designed by the researcher keeping his research topic and its objectives in mind. 4.5.1 Scales of Measurement There are four different scales of measurement; nominal, ordinal, interval and ratio. The rules used to assign numerals objects define the kind of scale and level of measurement. A brief account of each scaling type is given below, i. Nominal Scales Nominal scale is the simplest form of measurement. A variable measured on a nominal is one which is divided into two or more categories, for example, gender is categorised as male or female, a question as to whether a family owns a T.V. can be answered Yes or No. It is simply a sorting operation in which all individuals or units or answers can be placed in one category or another (i.e. the categories are exhaustive). The essential characteristic of a nominal scale is that in terms of a given variable, one individual is different from another and the categories are discriminate (i.e. the categories are mutually exclusive). This characteristic of classification if fundamental to all scales of measurement. Nominal scales that consist only two categories such as female-male, rich-poor, yes-no, are unique and are called dichotomous scales. ii. Ordinal Scales Ordinal scales have all the properties of a nominal scale, but, in addition, categories can be ordered along a continuum, in terms of a given criterion. Given three categories A, B and C, on an ordinal scale, one might be able to say, for e.g., that A is greater than B and B is greater than C. If numerals are assigned to ordinal scale categories,

the numerals serve only as ranks for ordering observations from least to most in terms of the characteristic measured and they do not indicate the distance between scale that organises observations in terms of categories such as high, medium and low or strongly agree, agree, not sure, disagree, and strong disagree. iii. Interval Scales Interval scales incorporate all the properties of nominal and ordinal scales and in addition, indicate the distance or interval between the categories. In formal terms one can say not only that A is greater than B and B is greater than C but also that (A-B)=(B-C) or (AC)=(A-B)+(B-C). Examples of interval scale include age, income and investments. However, an interval scale is one where there is no absolute zero point. It can be placed anywhere along a continuume.g., the age can be between 20 to 60 years and need not necessarily start from 0 years. This makes ratio comparison, that A is twice that of B or so wrong. iv. Ratio Scales A special form of interval scale is the ratio scale which differs in that it has a true zero point or a point at which the characteristic that is measured is presumed to be absent. Examples of ratio scales include, weight, length, income, expenditure and others. In each there is a concept of zero income, zero weight, etc. Since ratio scales represent a refinement of interval scales, generally these scales are not distinguished and both the terms are used inter-changeably. Nominal scales that consist only two categories such as femalemale, rich-poor, aware - unaware, etc. are unique. Such dichotomous nominal scales are important to researchers because the numerical labels for the two scale categories can be treated as though they are of interval scale value.

Each of the above four types of scales have a unique method of measurement. Both nominal and ordinal scales consist of discrete number of categories to which numbers are assigned. Thus, a variable such as number of families owning a car or CTV can only take values of 0, 1, 2 3 4 etc. It cannot have values such as 1.5 or 2.5 as the units are integers and indivisible. But interval and ratio scales take any value between two integers, as the variables are continuous. For example, given any ages however close, it is possible to find a third which lies in between. Interval and ratio scales are superior to normal and ordinal scales and a wealth of statistical tools can be employed in their analysis. The different statistical tools are related to these different scales of measurement, in that there is usually a correspondence between mathematical assumptions of the statistical tool and the assumptions of the scale of measurement. Care must be always taken to match the tools used with the scale of measaremerit of variables and to use a method which implies a higher scale measurement than the variable allows. 4.5.2 below. i. Paired comparisons As the name suggests, paired comparisons require the Data collection methods for qualitative data

There are several types of data collection methods. They are dealt

respondent to choose one of a pair of stimuli that "has more of," "dominates,"' "precedes," "wins over," or "exceeds" other with respect to some designated property of interest. If, for example, five dishwashing detergent brands are to be compared for "gentleness on the hands," a full set of paired comparisons (if order of presentation is

not considered) would involve (5 * 4)/2, or 10, paired comparisons. In the columns the preferred brands are taken. If the column brand is preferred over the row candidate brand, 1 is entered in the column cell and 0 is written in the row brand. Then column totals are made which indicate the relative strength the different column brands. Refer the chart given below: the diagonal cells carry * marks as these are self comparisons and no ranking is involved. Brand D is the most preferred one, followed by B and so on. Preferred Brands Candidate Brands A B C D E Total * 0 1 0 1 2 1 * 1 0 1 3 0 0 * 0 1 1 1 1 1 * 1 4 0 0 0 0 * 0 A B C D E

ii. Ranking methods Ranking procedures require the respondent to order stimuli with respect to some designated property of interest. For example, instead of using the paired comparison technique for determining the perceived order of the dishwashing detergents with respect to "gentleness on the hands," each respondent might have been asked to rank the detergents with respect to that property. When there are n items to be ranked , n-1 ranks have to be made, with the* last one taking the nth place.

iii. Ordered - category sorting Ordered category sorting involves putting a given number of stimuli in sets of different levels of preference or importance or so. When there are hundred items simple ranking will be a Herculean task. But putting them in 10 categories of 10 each in the order of importance is relatively easier.. For example, 15 varieties of dishwashing detergents representing the stimuli set, the respondent might be asked to sort them into three ordered categories of five each; (1) highly gentle on the hands, (2) moderately gentle, and (3) not gentle at all. iv. Rating methods Rating methods, as used in both marketing research and the behavioral sciences, represent one of the most popular and easily applied data collection methods. Rating method can take several forms: (1) numerical, (2) graphic, or (3) verbal. Often two or more of these formats appear together, as illustrated in the Figure. Verbal rating could be as follows: Excellent, Very Good, Good. Above Average, Average, Below Average, Poor, Very Poor and Worst. Equivalent numerical rating could be 9,8,7,6,5,4,3,2 and l.As shown in the figure, the respondent is given both verbal descriptions of the degree of "Excellent to Worst." and a series of integers (1 through 9) In the Figure the respondent need only to check the category appropriate for expressing his feelings about some attitude statement regarding dishwashing detergents. The respondent would be then asked to circle the number associated with the descriptive statement that comes closest to how he feels about the quality of the, brand(s) being rated. Figure: Verbal - Numeric Rating Scale. Excellent 9 Very Good 8 7 Good Above Average 6 5 Average Below Average 4 3 Poor Very Poor 2 1 Worst

v. Fractionation Fractionation is a type of quantitative-judgment procedure in which the respondent is given two stimuli at a time (e.g., a standard dishwashing detergent and a test brand) and asked to give some numerical estimate of the .ratio between them, with respect to some attributes, such as "harshness on the kinds." vi. Constant sum Constant - sum methods primarily because of their simplicity and ease of instructions, have become quite popular in marketing research. In constant - sum methods the respondent is given some manner of points -typically 10 to 100 - and asked to distribute them over the alternatives in a way that reflects their relative magnitude regarding some additional characteristics. 4.5.3. Methods of scale construction There are different types of scale construction techniques. To study stimuli such as effectiveness advertisement copies, leadership styles, different brands of a product, different competing service providers, etc. a particular scaling technique and for studying subjects like salesmen effectiveness, employees' satisfaction, customer awareness levels, etc. another scaling technique are needed, the former scaling is called stimuli centred scale and the latter, is called subject centred scale. 4.5.3.1 Stimuli centred scale ; Semantic Differential scale Essentially, semantic differential scale enables the researchers to probe into both the directions and the intensity of respondents* attitude towards such concepts as corporate image, advertising image,

brand or service image, and so on. One way this is done is to ask the respondent to describe the company by means of ratings on a set of bipolar adjectives. For measuring the Corporate Image of certain entities, we may use the following scale. 3 2 1 0 -1 -2 -3

Innovative Customer focused Dynamic Futuristic

Dull Shifting focus Static Immersed in past

On each of the adjective continuum, innovative - dull, customer focused - shifting focus, dynamic - static and futuristic - immersed in past, the chosen company is rated. If it is a valuable firm, it checks on the extreme left side of the continuum, else on the right of the continuum. 4.5.3.2. Subject centred scale : Likert Summated Scale

The summated scale was originally proposed by Rensis Likert, a psychologist. To illustrate, assume that the researcher wishes to scale some characteristics, such as the public's attitude toward advertising. In applying the Likert summated scale technique, the following steps are typically carried out: 1. The researcher assembles a large number (e.g., 75 to 100) of statements advertising. concerning the public's sentiments toward

2.

Each of the test items is classified by the researcher as generally "favourable" or "unfavourable" with regard to the attitude under study. No attempt is made to scale the items; however, a pretest is conducted that involves full set of statements and a limited sample of respondents.

3.

In the pretest the respondent indicates approval (or not) with every item, checking one of the following direction - intensity descriptors: a. b. c. d. Strongly approve Approve Undecided Disapprove

e. Strongly disapprove 4. Each response is given a numerical weight (e.g., 5, 4, 3, 2, and ] from strongly approve through strongly disapprove for positive statements and the scores are reversed for negative statements. 5. The individual's total - attitude score is represented by the algebraic summation of weights associated with the items checked. In the scoring process, weights are assigned such that the direction of the attitude - favourable to unfavourable - is consistent with items. For example, if a 5 were assigned to "strongly approve" for favourable items, a 5 should be assigned to "strongly disapprove" for unfavourable items. 6. After seeing the results of the pretest, the analyst selects only those items that appear to discriminate well through high and low total scores. Then, the main difference on each specific item are compared between these high and low groups (excluding the middle 50% of subjects).

7.

The 10 to 15 items finally selected are those that have discriminated "best" (i.e., exhibited the greatest difference in mean values) between high versus low total scorers in the pretest.

8. 4.6

Steps 3 to 5 are then repeated in the main study. PRE-TESTING The tool or instrument of data collection namely the schedule or

the questionnaire should be pre-tested before adopted for data collection on the study. Pre-testing simply means, testing the validity, reliability, practicability and sensitivity of the tool before it is used for actual data collection. The only way to gain assurance that questions 'ire unambiguous is to try them on a selected small group of prospective respondents. 4.6.1 Process of Pre-testing Pre-testing can be done in parts. Different sub-parts in the main part of the questionnaire / schedule can be differently pre-tested. So a series of small pre-test on different units of the tool can be done. A full scale pre-test of the whole tool can be done if needed finally or in lieu of the series of pre-tests is small bits. Pre-testing must be done on a sample that is representatives of the population. May be 10 to 12 respondents for pre-testing are good.

4.6.2 Importance of pre-testing The following are the objects of pre-test i) To detect discrepancies in the tool and rectify the same. This is needed to find where the shoe bites and making amends for the same. ii) To detect the difficulties encountered by the respondents while filling up the questionnaire / schedule and make remedies for the same. iii) To detect possible misunderstood, un-understood aspects of the tool and rectify the same. iv) The sequence of questions is better ordered in the light of feedback received. v) To get now insights into the problem based on responses received through pre-test and incorporate them in the tool and thereby enriching the tool. vi) To take note of flabby parts in the tool and remove them to make the tool slim and fit. vii) To re-size the tool based on time taken for filling up the questionnaire. The tool is thus right sized. This is akin to 'trial; of coat-suit before final stitching is made. Goode and Hatt indicated the following as signs or symptoms of defective schedule / questionnaire which may be seen during pretesting: (i) lack of proper order in the responses, (ii) all are none responses, (iii) large number of "do not know" or "do not understand" answers, (iv) many qualified answers or irrelevant opinions, (v) high

proportion of refusals to answer, (vi) high degree of variance in answers when the questions order is changed, (vii) similar type of people giving different types of replies, (viii) answers differing known answers, (ix) in case of questionnaire refusal to answer it, (x) marginal comments of interviewers or respondents concerning particular questions. 4.6.3. Advantage of Pre-testing The advantages of pre-testing are as follows: (i) The strengths and weaknesses of the tool are known beforehand. (ii) The structure and contents of the tool are refined in the light of weaknesses spotted. (iii) Measurement problems like offensive questions, intrusive or invasive questions get rectified. (iv) Effectiveness of the tool the like. (v) An estimate of resources needed in terms of non-hours is enhanced by rightly working

certain questions, removing confusion causing questions and

needed, enumerates needed, etc. to conduct the final survey can be made. 4.7 PILOT STUDY Pilot study is a theatrical dress rehearsal of the final field study. Pilot study sometimes is done to pre-test the tool and at other times after pre-test of tool to have a test-run of the field study. A pilot study is done on a small sample of respondents of the chosen population.

The purpose of pilot study is different from pre-testing. Pilot study is a kind of pre-testing of the "how" of conducting the field survey of study. The administration of the tool on the sample respondents is thus rehearsed. Based on the quality and quantity of responses got, alternatives terms of administration are decided upon. Pilot study should not be confused with 'research reconnaissance' measure done at the time of problem definition to get into grips with the problem or issue studied. 4.8 OBTAINING SECONDARY DATA Secondary data are obtained from personal documents and public documents. 4.8.1 Personal Documents These documents are recorded by the individuals. An individual may record his views and thoughts about various problems and without knowing for these documents at a latter data so formed a subject or source of study. Kinds of Personal Documents : Personal documents may be categorised or divided under the following heads for the convenience of the study; (i) Life History, (ii) Dairy, (iii) Letters, and (iv) Memoirs. i) Life History Life history, generally speaking contains all kinds of

biographical , material, from the point of view of personal documents only an autobiography which contains description and views about social and personal events is a life history. It may be further classified under the following three sub-heads: Spontaneous Autobiography, Voluntary Autobiography of self record, and Compiled life history.

ii) Diaries Many people keep diaries in which they record the daily events of their life and their feelings and reactions relating to those events. Some of the diaries are also published later on. Diaries are the most important source of knowing the life history of a person. If they have been written continuously over long periods. iii) Letters Letters also provide useful and reliable material on many social problems. They throw light upon more intimate aspects of an event, and clarity the stand taken by a person regarding it. They are helpful in giving an idea of the attitudes of a person and the trend of his mind. The validity of letters is beyond all doubt and they should be accepted as prima facie proof of the attitude of the writer. In such social problems as love, marriage or divorce the letters can supply much revealing information. iv) Memoirs Some people write memoirs of their travels, important events of their life and other significant phenomena that they come across. These memoirs provide useful material in the study of many a social phenomena. Memoirs are different from dairies in the sense that they describe only some events and are more elaborate than the dairy. Memoirs of travelers have provided us with useful information regarding the language, social customs, religious faiths, culture and many other social aspects of the people they visited. 4.8.2 Public Documents Public documents are quite different from personal documents. They deal with the matters of different interest. Public' documents

may be divided into the following two categories: (i) Unpublished Records, (ii) Published Records. Unpublished Records: Unpublished records give matters of public interest not available to people in published form. Everybody cannot have access to them. Proceedings of the meetings, noting on the files and memoranda etc., form the category of unpublished records. It is said that these records are reliable. Since there is no fear of their being made public the writers give out their views clearly. Published Records: Published records are available to people for investigation and perusal. Survey, reports, report of survey enquiries and such other documents fall under this category. The data contained in these documents are considered by some people as quite reliable because the collecting agency knows that it shall be difficult to test, while others are of the view that if the data are to be published the collecting or publishing agency does some window dressing, as a result of which the accuracy is sometimes postulated. Now most of the information that is available to people and researchers in regard to social problems is to be found in form of reports. The reports published by Government are considered as more dependable. On the other hand some people think that the reports that are published by certain individuals and agencies are more dependable and reliable. Journals and Magazines: Journals and magazines are

important public documents including a wide variety of information which .can be usefully utilised in social research. Most of these information are very much reliable. Letters to the editors published in

various

magazines

and

journals

are

an

important

source

of

information. Newspapers: Newspapers publish news, discussion on

contemporary issues, reports of meetings and conferences, essays and articles on living controversies and the letters of the readers to the editors. All this is an important source of formation for different kinds of social research. Other Sources: Besides the above mentioned public

documents, film, television, radio and public speeches etc., are other important sources of information. They supply useful information about contemporary issues. The investigator, however, should be capable of sorting out the reliable material and distinguishing it from the unreliable material advanced by these sources. 4.8.3 Sources of Secondary Data There are two broad categories of secondary data - internal secondary data and external secondary data. i) Internal secondary data Internal (secondary) data refers to information that already exists within the company in which the research problem arises. For instance, in many companies, salesmen routinely record and report their sales. Examples of secondary data include records of sales, budgets, advertising and promotion expenditures, previous marketing research studies and similar reports. Use of such secondary data can help the marketing manager analyse the effect of the different elements of the marketing mix, develop a marketing plan, make Budget and sales territory allocations, and, in general, help in managerial decision making.

ii) External secondary data External (secondary) data refers to information which is collected by a source external to the firm (whose major purpose is not the solution of the particular research problem facing the firm). There are three major categories of external data; 1. 2. 3. 4.8.4 Government sources and publications Business reference sources Commercial agencies Advantages of Documentary Sources of Data Following merits are usually claimed for using documentary source. (i) Provides an insight into total situation. The purpose of use of available materials is to explore the nature of the data and the subjects to get an insight into the total situation. While looking for the data required by the researcher he may uncover many more available data than are often assumed to exists and hence contributes significantly to the unfolding of hidden information. (ii) Helps in the formulation of hypothesis. The use of

documentary sources sometimes, helps in the formulation of research hypothesis. While an investigator may have one or two hypotheses which he might have deduced from theory, the study of available materials may suggest further hypotheses. If a research idea or hypotheses can be formulated in such a manner that the available recorded material bears on the question, the use of such material becomes possible.

(iii)

Helps in testing the hypotheses. The available records may also help in testing the hypothesis.

(iv)

Provides supplementary information. Available documents may be used to supplement or to check information gathered specifically for the purposes of a given investigation. For example, if one has drawn in random sample of a small group in order to interview individuals, the accuracy of one's sample could be checked by comparing socio-economic data of the sample, like income, education standard caste, family size etc., with the same data of the most recent census or with available data in local Government offices.

4.8.5 Disadvantages of Secondary Sources of Data The following are the demerits of secondary data i) Collected for a specific purpose Data are often collected with a specific purpose in mind, a purpose that may produce deliberate or unintentional bias. Thus, secondary sources must be evaluated carefully. The fact that secondary data were collected originally for particular purposes may produce other problems. Category definitions, particular measures or treatment effects may not be the most appropriate for the purpose at hand. ii) Old data Secondary data are by definition, old data. Thus, the data may not be particularly timely for same purposes.

iii)

Aggregation of data in Inappropriate Unit Seldom are secondary data available at the individual

observation level. This means that the data are aggregated in some form, and the unit of aggregation may be inappropriate for a particular purpose. iv) Authenticity The authenticity of same secondary sources of data is doubtful. v) Context change Secondary data refer to a given situation. As situations change, the data lose their contextual validity. QUESTIONS 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Explain the nature and features of primary data. Explain the nature and features of secondary data. Distinguish primary data from secondary data, What precautions are needed in using secondary data? Give an account of different methods of data collection. What is analytical method of data collection? What is case study? How is it a method of data collection? What are the different kinds of observation? How are the relevant as data collection method? What is survey method? Give the different types of surveys. Explain interview what are its importance and purpose as data collection method? Discuss the features of different types of interview. How do you prepare yourself as a researcher for conducting an interview? Write about the features of the effective interview technique and limitations of the interview technique.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

What is a schedule? Explain its kinds and the essentials of a goal schedule. How do you prepare a schedule? What are the steps involved? What is a questionnaire? Explain different types of questionnaire as a tool for data collection. Give the format of a questionnaire for a study of your choice. Examine the factors that influence response to questionnaire. Bring out the merits and demerits of schedule questionnaire. What is scaling? State the different scales and their features and importance. Explain different data collection practices using scaling technique. How do you construct a scale? Explain with an example. How is Likert scale is constructed? How is semantic different scale constructed? What is pre-testing? Explain its need and merits.

REFERENCES 1. 2. 3. 4. 5. 6. 7. Research Materials - P.Saravanavel Elements of Educational Research - S.P.Sukhia Research for Marketing Decisions - Green & Tull Marketing Research - Sangeetha Agarwal Research Methdology - Basetia and Sharma Encyclopadie of Modern Education Research - S.K. Sharma Research Methodology - C.R. Kothari

UNIT V

PROCESSING AND ANALYSING DATA

In this unit the meaning, importance and process of data processing and analysis editing, coding, classification, tabulation, diagramming, presented. INTRODUCTION Processing of data is the most important step in research. Lot of statistical treatment is done on the data. After such analysis interpretation of the statistics is made to draw inferences about the population. 5.1 MEANING OF PROCESSING AND ANALYSIS OF DATA The data collected need to be processed and analyzed. Processing of data means editing, coding, classification, tabulation, charting and diagrammatic presentation of data. Analysis of data means computing appropriate statistical measures such as measures of central tendency, measures of dispersion, measures of association, computing test statistics and drawing inferences from all the above. 5.2 IMPORTANCE OF DATA PROCESSING AND ANALYSIS To distil the essence of information from the mass of data, processing and analysis of data are essential. Data processing and analysis is the culmination of the long process of research design, survey and data collection. From the mass of data meaning must be churned out. This needs in the first place editing to do away with unnecessary data, irrelevant data and the like. Thus some data reduction is effected. Further reduction of data is done through coding which replaces like ones by a standard code. After coding, interpretation and guidelines therefore, etc. are

classification is done which groups data according to common features found or established. This is another data reduction purpose. Tabulation then follows which involves, presenting classified data in the form of tables. Each table gives summary data on an issue our related issues. A table presents essential data on a particular topic and enables making sense out of the figures. All these above are certain aspects of data processing. The purpose is to figure out essential facts from a vast mass of data. This facilitates analysis, the next step. Analysis involves computing i) descriptive statistics or

parameters like measures of central tendency, ii) associative statistics such as correlation and regression and iii) test statistics such as Z values, Y values, 'F1 values, chi-square values, 'U' values, and so on and iv) finding out clusters, principal components, etc. Without these analyses, no inference, no conclusion and no recommendation could be offered. So, description, association, hypotheses testing and inference are possible only when the basic aspect of processing and relevant analyses are made. Overall understanding of the problems, issues, phenomenon and events studied is facilitated by data processing and processing. Description, implantation and prediction of phenomena, problems and issues studied are facilitated only by processing and analysis. A major part of main body of any research work has jls bases in processing and analysis. Hence the importance of data processing and analysis. 5.3 PROCESSING OF DATA Data processing is concerned with editing, coding, classifying, tabulating and charting and diagramming data. The essence of data

processing is data reduction. Data reduction involves winnowing out the irrelevant from the relevant data and establishing order from chaos and giving shape to a mass of data. 5.3.1 Editing Editing is the first step in data processing. Editing is the process of examining the data collected in questionnaires/schedules to detect errors and omissions and to see that they are corrected and the schedules are ready for tabulation. When the whole data collection is over a final and a thorough check up is made. Milred B. Parten in his book points out that the editor is responsible for seeing that the data are (1) accurate as possible, (2) consistent with other facts secured, (3) uniformly entered, (4) as complete as possible, (5) acceptable for tabulation and arranged to facilitate coding tabulation. The editor must not destroy data, make illegible the original data, make the editing work in a different ink, all changes must be initialed and dated. There are different types of editing. Editing for quality, Editing for tabulation, Field editing and Central editing are four types of editing. Editing for quality asks the following questions: are the data forms complete, are the data free of bias, are the recordings free of errors, are the inconsistencies in responses within limits, are there evidences to show dishonesty of enumerators or interviewers and are there any wanton manipulation of data. Editing for tabulation does certain accepted modification to data or even rejecting certain pieces of data in order to facilitate tabulation.

or instance, extremely high or low value data item may be ignored or bracketed with suitable class interval. Field Editing is done by the enumerator. The schedule filled up by the enumerator or the respondent might have some abbreviated writings, illegible writings and the like. These are rectified by the enumerator. This should be done soon after the enumeration or interview before the loss of memory. The field editing should not extend to giving some guess data to fill up omissions. Central Editing is done by the researcher after getting all schedules or questionnaires or forms from the enumerators or respondents. Obvious errors can be corrected. For missed data or information, the editor may substitute data or information by reviewing information provided by likely placed other respondents." A definite inappropriate answer is removed and "no answer" is entered when reasonable attempts to get the appropriate answer fail to produce results. Editors must keep in view the following points while performing their work: (i) They should be familiar with instructions given to the interviewers and coders as well as with the editing instructions supplied to them for the purpose, (ii) While crossing out an original entry for one reason or another, they should just draw a single line on it so that the same may remain legible, (iii) They must make entries (if any) on the form in some distinctive colour and that too in a standardized form, (iv) They should initial all answers which they change or supply, (v) Editor's initials and the data of editing should be placed on each completed form or schedule.

5.3.2 Coding Coding is necessary for efficient analysis and through it the several replies may be reduced to a small number of classes which contain the critical information required for analysis. Coding decisions should usually be taken at the designing stage of the questionnaire. This makes it possible to pre-code the questionnaire choices and which in turn is helpful for computer tabulation as one can straight forward key punch from the original questionnaires. But in case of hand coding some standard method may be used. One such standard method is to code in the margin with a coloured pencil. The other method can be to transcribe the data from the questionnaire to a coding sheet. Whatever method is adopted, one should see that coding errors are altogether eliminated or reduced to the minimum level. * Coding is the process/operation by which data/responses are organized into classes/categories and numerals or other symbols are given to each item according to the class in which it falls. In other words, coding involves two important operations; (a) deciding the categories to be used and (b) allocating individual answers to them. These categories should be appropriate to the research problem, exhaustive of the data, mutually exclusive and uni - directional Since the coding eliminates much of information in the raw data, it is important that researchers design category sets carefully in order to utilize the available data more fully. The study of the responses is the first step in coding. In the case of pressing - coded questions, coding begins at the preparation of interview schedules. Secondly, coding frame is developed by listing the possible answers to each question and assigning code numbers or symbols to each of them which are the indicators used for coding. The coding frame is an outline of what is coded and how it is to be coded. That is, a coding frame is an outline of what is coded and how it is to

be coded. That is, coding frame is a set of explicit rules and conventions that are used to base classification of observations variable into values which are which are transformed into numbers. Thirdly, after preparing the sample frame the gradual process of fitting the answers to the questions must be begun. Lastly, transcription is undertaken i.e., transferring of the information from the schedules to a separate sheet called transcription sheet. Transcription sheet is a large summary sheet which contain the answer/codes of all the respondents. Transcription may not be necessary when only simple tables are required and the number of respondents are few. 5.3.3 Classification Classification or categorization is the process of grouping the statistical data under various understandable homogeneous groups for the purpose of convenient interpretation. A uniformity of attributes is the basic criterion for classification; and the grouping of data is made according to similarity. Classification becomes necessary when there is a diversity in the data collected for meaningless for meaningful presentation and analysis. However, it is meaningless in respect of homogeneous data. A good classification should have the characteristics of clarity, homogeneity, equality of scale, purposefulness and accuracy. Objectives of Classification. Objectives of Classification are below: (a) Through classification. The complex scattered and haphazard data is organized into concise, logical and intelligible form. (b) Through classification, it is possible to make the characteristics of similarities and dis - similarities clear.

(c) (d)

Through classification, comparative studies is possible. Through classification, understanding of the significance is made easier and thereby good deal of human energy is saved.

(e)

Underlying unity amongst different items is made clear and expressed.

(f)

Data is so arranged that analysis and generalization becomes possible. Types of classification. Classification is of two types, viz.,

quantitative classification, which is on the basis of variables or quantity; and qualitative classification (classification according to attributes). The former is the way of, grouping the variables, say, quantifying the variables in cohesive groups, while the latter groups the data on the basis of attributes or qualities. Again, it may be multiple classification or dichotomous classification. The former is the way of making many (more than two) groups on the basis of some quality or attributes while the latter is the classification into two groups on the basis of presence or absence of a certain quality. Grouping the workers of a factory under various income (class intervals) groups come under the multiple classification; and making two groups into skilled workers and unskilled workers is the dichotomous classification. The tabular form of such classification is known as statistical series, which may be inclusive or exclusive. 5.3.4 Tabulation Tabulation is the process of summarizing raw data and displaying* it in compact form for further analysis. Therefore, preparing tables is a very important step. Tabulation may be by hand, mechanical, or electronic. The choice is made largely on the basis of

the size and type of study, alternative costs, time pressures, and the availability of computers, and computer programmes. If the number of questionnaire is small, and their length short, hand tabulation is quite satisfactory. Table may be divided into: (i) Frequency tables, (ii) Response tables, (iii) Contingency tables, (iv) Uni-variate tables, (v) Bi-variate tables, (vi) Statistical table and (vii) Time series tables. Generally a research table has the following parts: (a) table number, (b) title of the table, (c) caption (d) stub (row heading), (e) body, (f) head note, (g) foot note. As a general rule the following steps are necessary in the preparation of table: (i) Title of table: The table should be first given a brief, simple and clear title which may express the basis of classification. (ii) Columns and rows: Each table should be prepared in just adequate number of columns and rows. (iii) Captions and stubs: The columns and rows should be given simple and clear captions and stubs. (iv) Ruling: Columns and rows should be divided by means of thin or thick rulings. (v) Arrangement of items; Comparable figures should be arranged side by side. (vi) Deviations: These should be arranged in the column near the original data so that their presence may easily be noted.

(vii)

Size

of

columns:

This

should

be

according

to

the

requirement. (viii) Arrangements of items: This should be according to the problem. (ix) Special emphasis: This can be done by writing important data in bold or special letters. (x) Unit of measurement: The unit should be noted below the lines. (xi) (xii) Approximation: This should also be noted below the title. Foot - notes: These may be given below the table.

(xiii) Total: Totals of each column and grand total should be in one line. (xiv) Source : Source of data must be given. For primary data, write primary data. It is always necessary to present facts in tabular form if they can be presented more simply in the body of the text. Tabular presentation enables the reader to follow quickly than textual presentation. A table should not merely repeat information covered in the text. The same information should not, of course be presented in tabular form and graphical form. Smaller and simpler tables may be presented in the text while the large and complex table may be placed at the end of the chapter or report.

5.3.5

Diagrams: Diagrams are charts and graphs used to present

data. These facilitate getting the attention of the reader more. These help presenting data more effectively. Creative presentation of data is possible. Interpretation is effectively done, 5.3.5.1 Charts A chart is a diagrammatic form of data presentation. Bar charts, rectangles, squares and circles can be used to present data. Bar charts are uni-dimensional, while rectangular, squares and circles are two-dimensional. Bar Charts : Bar charts refers to bars employed to represent data. Vertical and holrizontal bars can be used. There are simple bar charts, multiple bar charts, sub-divided bar charts, broken bar charts and percentage bar charts. Simple bar charts : Simple bar charts employ bars to represent data. The length of! the bars runs proportional to the value of data. Suppose, in a month, in a branch of a life insurance firm, 360 endowment policies, 480 whole life policies and 120 annuities were sold. These data can be expressed in bars as below: (Vertical bars are used). (1) Endowment Policies (2) Whole lie Policies (3) Annuities

Multiple Bar Charts : Multiple bar charts employ blocks of bars, each block of bars represent an entity and bars in a block represent values of variables studied. Suppose the number of endowment, while life and annuity policies sold by two branches are: 120, 360 and 240 for one branch and 480, 240 and 360 for another branch. Multiple bar charts are used as follows:

Subdivided Bar Charts : The above data can be given in subdivided bar charts as well. Here, one bar for Branch-I and another for BranchII will be, used. The length of the bar will correspond to total number of policies and the bar -will be divided to represent different classes of policies sold. Vertical or horizontal bars can be used. Here horizontal bars are used as below:

Broken Bar Charts : Broken bar charts are employed when few quantities are too large and the rest are small. In such cases, the bars to represent the large quantities are broken, so that economic use of space and better look of the presentation are possible. Take the case: One branch of LIC sold 100 policies in a month, another 500 policies and another one 5000 policies. A better way of chart presentation is the broken bar chart form as shown below:

Pie Chart : Pie-chart is one of the most important charts in twodimensional diagram. It is a representation of various data on the basis of different character When a circle is sub-divided into various sectors, it is called pie-chart. The pie chart shows the relationship of parts to the whole. The following steps are involved in the construction of pie-chart i) ii) Draw a circle of any radius Compute the angle representing each sector by the formula The angle of sector A = Value of sector A Total Value iii) Arrange the angles in ascending or descending order of magnitude

X 360

iv)

Use shades or colours to distinguish sectors.

The following example illustrates the procedure of constructing piecharts. Illustration From the following data about the percentage distribution of different categories of insurance policies sold by a general insurance company in a particular year, draw a pie-chart. Type of policies Fire Marine Mediclaim Motor-vehicle Others Percentage share 30% 18% 28% 14% 10%

Solution Calculating for the preparation of pie-chart Type of polices Fire Marine Mediclaim Motor-vehicle Others Total Percentage 30 18 28 14 10 100 Angle (30/100) x 360 -108 (1 8/100) x 360- 65 (28/100) x 360 -101 (14/100) x 360= 50 (10/100) x 360= 36 360

5.3.5.2

Graphs

The method of presenting numerical data in visual form is called graph, A graph gives relationship between two variables by means of either a curve or a straight line. Graphs may be divided into two categories. (1) Graphs of Time Series and (2) Graphs of Frequency Distribution. In graphs of time series one of the factors is time and other or others is / are the study factors. Graphs on frequency show the distribution of by income, age, etc. of executives and so on. Frequency distribution related to discrete and continuous series can be well drawn in a graph. The methods for drawing graphs for frequency distribution are as follows:

i) ii) iii) iv)

Histogram Frequency polygon Frequency curve Ogives or cumulative frequency curve

5.3.5.2.1

Histogram

The most commonly used graph for translating frequency distributions is : the histogram. It is a graph that represents the class frequencies in a frequency .distribution by vertical rectangles. It is constructed on a pair of coordinate axes. The class frequencies are plotted along the vertical axis and class intervals along the horizontal axis. The vertical axis (Y axis) must start with zero. The horizontal axis may start with any convenient number. For each class interval, we construct a rectangle of height proportionate to frequency of that class and width of whose base equals the width of the classinterval. In this way we get a series of adjoining rectangles. The combination of these rectangles is called histogram. Histogram is two dimensional. Histogram are of two types - (a) Histogram for frequency distributions having equal class-intervals and (b) Histogram for frequency distribution having unequal classintervals. i) Construction of Histogram Equal Class Intervals

Illustration : Draw a histogram from the following data Age 0-10 10-20 20-30 30-40 40-50 No. of people in a particular place 7 12 15 17 20

50-60 60-70 70-80 80-90

18 14 10 4

Solution : The class-intervals are taken on the x-axis and bars

proportional to frequencies are drawn on respective class-intervals. The graph below gives a histogram.

ii)

Construction of Histogram - Unequal Class Intervals : For the construction of histogram, frequencies are to be

adjusted in case of unequal class-intervals. Suppose the width of the one class is 3 times the width of the lowest class then its frequency should be divided by 3 for making adjustment. If the width of any

other class is 4 times compared to the width of the lowest class then its frequency should be divided, by 4 and so on. Illustration : The following frequency distribution shows the

ages of insurancers in a particular place. Age 8-12 12-16 16-20 20-28 28-40 40-56 No. of Insurances 9 15 18 20 15 12

Construct a histogram Solution : Construction of histogram Here, the last 3 classes are of unequal width, therefore three class-frequencies need adjustment. The frequencies of these 3 classes are to be adjusted on the basis of classes of equal width. The procedure of adjusting the frequencies is given below: Class Interval 4-8 8-12 12-16 16-20 Frequency 3 9 15 18 Adjusted Frequency

20-28 28-40 40-56

20 15 12

20/2=10 15/3 = 5 12/4 = 3

Histogram is drawn on the basis of adjusted frequencies. The following gives the histograph.

5.3.5.2.2

Frequency Polygon

A frequency polygon is a curve representing a frequency distribution. It is an effective graph for comparing two or more frequency distributions. A frequency polygon is constructed from a histogram by joining the mid-points of the tops of various rectangles by straight lines. The frequency polygon is completed by adding two line segments, one at each extreme class in the histogram, which connects the class marks of these classes with value of zero in the classes that would be outside of these extreme classes.

By another method, first we plot the frequencies corresponding to midpoints to get the frequency polygon curve. Let us construct frequency polygon from the histogram of equal class interval dealt a little earlier. First we reproduce the histogram Fig. as given below. The midpoints of the tops of various rectangles are joined together and the resulting curve is extended in both directions to the base line.

The special advantage of frequency polygons is that they can be super-imposed to facilitate comparison between two or more frequency distributions.

5.3.5.2.3

The smoothed Frequency Curve

The smoothed frequency curve is drawn freehand through the various point of the frequency polygon in such a way that the area under the frequency curve is approximately the same as that under the frequency polygon. It is necessary to construct either histogram or frequency polygon to draw a smoothened frequency curve. The curve may be extended in both directions. The smoothened frequency curve is very important in statistical analysis because it represents the true distribution of population from which sample is drawn. The curve follows for the above polygon dealt above. The smoothed Frequency Curve.

5.3.5.2.4

Ogives or Cumulative Frequency Curves.

A frequency distribution presented in a cumulative from is known as cumulative frequency curve or Ogives. Ogives are of two types; (i) (ii) Less than Ogive. More than Ogive

While constructing ogives, the values of the variable are represented along the X-axis, while the cumulated frequencies are plotted on the Y-axis. The cumulated frequencies are plotted upon lower class limits for constructing More than Ogives. Whereas in the case of "Less than Ogives" cumulated frequencies are plotted upon limits. On joining these plotted points, we get the required ogives. Less than Ogive; In a less than ogive, cumulating of frequencies is done upwards. The process of cumulating the frequencies with the upper limits of classes is illustrated below: Illustration : Draw less than ogives from the following data. Age 0-10 10-20 20-30 30-40 No. of Insurancers 5 13 25 18

40-50 50-60 60-70

12 8 4

Solution : Calculating for less than Ogive Age (Upper class limits) 10 20 30 40 50 60 70 No. of Insurancers (Cumulative Frequencies) 5 18(5+13) 43 (18 + 25) 61 (43 + 18) 73 (61 + 12) 81 (73 + 8) 85 (81+4)

It is evident from the above table that 5 persons are insured below the age of 10; 61 persons insured below 40years and so on. The graph of such a cumulative frequency distribution gives a rising cure as has been depicted in diagram below.

The less than ogive slopes upwards from left to right as depicted in the figure.

More than Ogive: In a more than ogive, the cumulating of frequencies is done downwards. Here, we start the process of cumulating the frequencies with lower class limits.

Age (Upper class limits) 0 10 20 30 40 50 60

No. of Insurancers (Cumulative Frequencies) 85 (80 + 5) 80(67+13) 67 (42+25) 42 (24 + 18) 24 (12 + 12) 12 (4 + 8) 4

The above cumulated table shows that 42 insurancers are of age more than 30, 12 insurancers are of more than 50 years of age and so on. The graph of such cumulated frequencies gives a falling curve as has been depicted in the figure below.

Construction of More than Ogive

The uses of ogives are many. The main use of cumulated frequencies is in the construction of Lorenz curve which is employed for studying the distribution of income in a community. The intersection of the two ogives gives the median value of the distribution. 5.3.6 Analysis of Data Analysis of data involves computing certain statistics or parameters. There are different measures of statistics. These are presented below briefly. 5.3.6.1 Central Tendency Analysis The main objective of statistical analysis is to represent the data by one single value which shows the concentration of data at that particular value. Such a value is called the central value which

facilitates easy comparison between two or more series compared to loose data. In case direct comparison is made between the wages of workers of two factories, A and B employing 1000 workers each we would be unable to reach any conclusion. If each of these two distributions of wages of workers is represented by one figure, it could facilitate easy comparison. Thus various measures which are employed to measure this tendency are called measures of-central tendency. These measures of central tendency are also called 'averages'. Murray R Spiegel defines average as, "a value which is typical or representative of a set of data". Various types of averages are employed in statistical studies for summarizing the huge mass of data. There are six important averages mainly employed for studying the characteristics of statistical series. These averages are; i) ii) iii) iv) v) vi) 5.3.6.2 Arithmetic mean Median Mode Geometric Mean Harmonic mean (X) or (AM) (M) (Z) (GM) (HM)

Weighted Arithmetic Mean (WAM) Dispersion Analysis

Dispersion or variation refers to a measure of the degree of scatter of data from the central value in a series. According to W.F.L. Dick, "Dispersion or spread is the degree of the scatter or variation of the variables about a central value", A.L.Bowley defines dispersion, as "the measure of the variation of the items".

The following three series will clear the meaning of Dispersion

Series I: Series II: Series III:

2,4,4,16,36,63,120 21,25,36,38,41,41,43 35,35,35,35,35,35,35

In all the three series, the arithmetic mean is the same and is equal to 35. But the series are different. In the third series, there is no dispersion. In the second series there is some dispersion and in the first series the variables are very much scattered from the mean and it has more dispersion. In the 1st distribution median is 16 and mode is 4. In the 2nd distribution, median is 38 and mode in 41. For the 3rd the mean, mode and! median are the same as 35. Though all the three distributions have same mean,

distributions 1st and 2nd have dispersion. Dispersion is of two types, viz., Absolute Dispersion and Relative Dispersion When dispersion is expressed in terms of original units, i.e., units in which data are given, it is called absolute dispersion. Range, Inter-quartile range, Quartile Deviation, Mean Deviation and Standard Deviation are measures of absolute dispersion. When dispersion is expressed as an abstract number such as a ratio or a percentage, it is called relative dispersion. It is also called as co-efficient of dispersion because co-efficient is a pure number which is independent of the units of measurement. The following are the

main measures: Co-efficient of Range, Co-efficient of Quartile Deviation, Co-efficient of Mean Deviation and Co-efficient of Variation 5.3.6.3 Association Analysis

Correlation and regression form part of association analysis. Correlation Co-efficient tells the direction and extent of relationship between the variables. When there is negative correlation, change in one will produce opposite change in the other. This will help in decision making. If advertisement spending by firms and number of cars sold have high positive correlation, it implies that by spending more an advertisement, sales can be increased. Correlation, thus, has great business decision - support orientation. Correlation is used in forecasting. In production, investment, finance, marketing and personnel management areas. Correlation analysis is of great use. Regression measures the influence of one variable over other. Ya Lun Chow defines that "Regression analysis attempts to establish the nature of relationship between variables; that is to study the functional relationship between the variables and thereby provide a mechanism for prediction or forecasting". That is, the influence of independent variable over dependant variable is studied through regression. The impact of climatic factors on form productivity and in consumption pattern of people, the impact of premium on demand for insurance policies, etc., speak about regression 5.3.6.4 Multi-variate analysis

Thus for a single or two variables is/are involved. When more number of variables are involved, multi-variate analysis is needed. Multiple regression, multiple correlation, multiple discriminant analysis, analysis of variance (ANOVA), multi-variable analysis of variance (MANOVA) and canonical analysis are important ones.

i)

Multiple regression analysis: This analysis is adopted when the researcher has one dependent variable which is presumed to be a function of two or more independent variables.

ii)

Multiple discriminant analysis: This analysis is appropriate when the researcher has a single independent variable that cannot be measured, but can be classified into two or more groups on the basis of certain attribute. The object of this analysis happens to be to predict an entity's possibility of belonging to a particular group based on several predictor variables.

iii)

Multivariate analysis of variance (or multi-ANOVA); This analysis is an extension of two-way ANOVA, wherein the ratio of among group variance to within group variance is worked out on a set of variables.

iv)

Canonical analysis: This analysis can be used in case of both measurable and non-measurable variables for the purpose of simultaneously predicting a set of dependent variables from their joint covariance with a set of independent variables.

v)

Inferential analysis: This is concerned with the various tests of significance for testing hypotheses in order to determine with what validity data can be said to indicate some conclusion or conclusions. It is also concerned with the estimation of population values. It is mainly on the basis of inferential analysis that the task of interpretation (i.e., the task of drawing inferences and conclusions) is performed.

5.3.6.5 Inferential analysis Inferential analysis is concerned with testing the significance of mean, difference between means, for proposition, for difference between propositions, for variance, for difference between variance, significance difference among means of three or more populations, significance difference among means of three or more groups, significance of correlation and regression coefficients, etc, come under this category. This is the most important analysis. (Note: You are advised to refer standard text books. Of course, certain fundamental ones were dealt in Chapter IV already). 5.4 PROCESS OF INTERPRETATION Interpretation means drawing inference from the collected facts and computed statistics or test statistics. The task of interpretation has two major aspects viz., (i) the effort to establish continuity in research through linking the results of given study with those of others, and (ii) the establishment of some explanatory concepts. "In one sense, interpretation is concerned with relationship within the collected data, partially overlapping analysis. Interpretation also extends beyond the data of the study to include the results of other research, theory and hypothses. Thus, interpretation is the devise through which the factors that seem to explain what has been observed by researchers in the course of the study can be better understood and it also provides a theoretical conception which can serve as a guide for further research. 5.4.1 Process of interpretation The of interpretation is not an easy job, rather it requires a great skill and dexterity on the part of researcher. Interpretation is an art that one learns through practice and experience. The researcher may,

at

times,

seek

the

guidance

from

experts

for

accomplishing

interpretation. The technique of interpretation often involves the following steps: (i) Researchers must give reasonable explanations of the relations which he has found and he must interpret the lines of relationship in terms of the underlying process and must try to find out the thread of uniformity that lies under the surface layer of his diversified research findings. In fact, this is the technique of how generalization should be done and concepts be formulated. (ii) Extraneous information, if collected during the study, must be considered while interpreting the final results of research study, for it may prove to be a key factor in understanding the problem under consideration. (iii) It is advisable, before embarking upon final interpretation, to consult someone having insight into the study and who is frank and honest and will not hesitate to point out omissions and errors in logical argumentation. Such a consultation will result in correct interpretation and, thus, will enhance the utility of research results. (iv) Researchers must accomplish the task of interpretation only after considering all relevant factors affecting the problem to avoid false generalization. He must be in no hurry while interpreting results, for quite often the conclusions, which appear to be all right at the beginning, may not at all accurate.

5.4.2

Need for interpretation The need for interpretation can hardly be over-emphasized.

(i)

It

is

through

interpretation

that

the

researcher

can

understand the abstract principle that works beneath his findings. Through this he can link up his findings with those of other studies, having the same abstract principle, and thereby can predict about the concrete worlds of events. Fresh enquiries can test these predictions later on. This way of continuity in research can be maintained. (ii) Interpretation leads to the establishment of explanatory concepts that can serve as a guide for future research studies; it opens new avenues of intellectual adventure and stimulates the quest for more knowledge. (iii) Researchers can better appreciate only through

interpretation why his findings are what they are and can make others to understand the real significance of his research findings. (iv) The interpretation of the findings of exploratory research study often results into hypotheses for experimental research and as such interpretation is involved in the transition from exploratory to experimental research. Since an exploratory study does not have a hypothesis to start with, the findings of such study have to be interpreted on a post-factum basis in which case the interpretation is technically described as 'post factum' interpretation.

5.4.3: The (i) (ii) (iii) (iv)

Guidelines for making valid interpretations: following guidelines are useful in making good

interpretations, Don't interpret from a single or limited number of instances Don't overstress both positive and negative points Don't omit evidences contrary to one's opinion. Don't overlook important circumstances pertaining to

different phenomena studied. (v) (vi) Don't have any pre-dispositions, pre-conceived notions Don't attribute results to a single factor when other factors are equally important. (vii) Don't base your judgment on inaccurate instruments of measurement (viii) Don't forget to distinguish between material and less-

significant issues (ix) (x) (xi) (xii) (xiii) Don't make any false analogy Don't generalize from insufficient data Try to see the problem in right perspective Make provision for unstudied factors Try to see the differences between cause and effect

(xiv) (xv) (xvi) (xvii)

Recognize the limitations of evidences Base judgments on complete and accurate data Ensure consistency of information and inferences Trust statistical evidences more than verbal evidences

(xviii) Elicit diverse opinions, make appropriate induction based interpretations (xix) Make appropriate use of deductive technique of

interpretation. QUESTIONS : 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. State the meaning and scope of data processing. State the meaning and scope of data analysis. Explain the importance of data processing. Explain the importance of data analysis. Explain the meaning, need and types of editing. Present the meaning, need and types of coding. Present the meaning, need and types of classification. Discuss the meaning, need and types of tabulation. Present the meaning, need and types of diagrams. Explain the tools of analysis of data. What is interpretation? How is it done? How is the quality of interpretation ensured.

REFERENCES: 1. 2. 3. 4. 5. 6. 7. Research Materials - P.Saravanavel Elements of Educational Research - S.P.Sukhia Research for Marketing Decisions - Green & Tull Marketing Research - Sangeetha Agarwal Research Methdology - Basetia and Sharma Encyclopadie of Modern Education Research - S.K. Sharma Research Methodology - C.R. Kothari

UNIT VI REPORT WRITING

In this last unit of this paper, Report writing is presented. Role of research reports, types of research reports, contents of research reports, steps involved in drafting research reports, principles of report writing, referencing and criteria of evaluating research reverts and research findings are presented. Introduction The culmination of the hard and committed work of the researcher is the research report which embodies every bit of his work right from the origin of research idea, the design of his research, the execution and finally the findings and suggestions that emanate from the study. The research report is considered as a major component of the research work, because through this report the problem, the design, the analysis and findings are brought to the knowledge of the world. The report converts the work into a public asset from its earlier state of private asset. 6.1 ROLE OF RESEARCH REPORT The research report shows the readers the progress in knowledge made in the specific area or discipline. The report by bringing to light the new frontiers of knowledge enhances the society's intellectual well-being. The report by highlighting the design and methodology, runs as a fore-runner for future researchers in this or related area. The analyses and interpretations may give a boost to knowledge. The findings and suggestions take the readers into enlightenment from ignorance. Every research must endeavour to achieve this.

Research report is a record of the whole of every bit of the research work. This document is a reservoir of knowledge for current and future references and use to solve societal problems. Research report is the means through which communication of the entire work to the society is made. For other researchers, a documented research is a source of information and that a research report generates research interests. Research report propagates throughout the humanity or the globe. The role of a research report is best known in the absence of the same-Assume for a while, that no researcher gives out his research work in the form of a report. Then the research work is just like a lamp in the pot. When, it takes the form of report it is like a lamp on the hillock illuminating the surroundings. If a research report is not made, even the researcher may not be able tell his work at a future date. Thanks to human's potentials to forget. Such waste of efforts should never occur. If only a research report was made out, reinventing the wheel would not take place otherwise, same problem may be analyzed by different people at different places or in the same place at different times or at the same time. This is a greater waste of human energy. Thus a research report conserves energy that would otherwise would have been spent uselessly. 6.2 TYPES OF REPORTS Research reports are of several types. These popular report, administrative report and technical report. The first one is intended for laymen, the second one for project lenders or administrators and the last one for future researchers. There are other types as well, based on the type of research. Accordingly we have case report, survey report, etc. Also on the basis on of individual or group effort, we have, personal research report and committee research report. The former is more knowledge

undertaken by academic degree oriented researchers and latter by R & D related personnel. 6.3 CONTENTS OF RESEARCH REPORT A research report generally contains three aspects:

i) Preliminary Section, ii) Main Body and iii) Reference Section. These are briefly given below: 6.3.1 Preliminary Section The preliminary section deals with title, acknowledgement, etc. Title Page: The title of the research report usually bears the investigator's name, a statement as to the course for which the study has been required such Ph.D. M.Phil., etc., the date of submission it, and the name of the institution making that requirement. In reports of studies not undertaken for any course, the investigator's name, the institution he belongs to and the date of completion of the work is indicated. In a published thesis the latter information is substituted or supplemented by the name of the publishers and the date and place of publication. Acknowledgement Page : The acknowledgement page is largely one of courtesy in which the investigator acknowledges the guidance and assistance he has received in the development of the study. Acknowledgement may not refer to the guide so much as to others who may have aided in a special way. It is rightly said that good taste calls for acknowledgements to be expressed simply and tactfully. Preface or Foreword : Sometimes a preface or foreword of one or two pages long, follows the acknowledgement page, bearing some initial

remarks and perhaps a brief statement of the scope, aim and general character of the research. Table of Contents : A well-developed table of contents renders a good deal of assistance to a reader in choosing rapidly and judiciously what he should, subsequently, read carefully. It is usually desirable to include in it not only the chapter headings, but also the headings of the major subdivisions of the chapters. Sometimes the topics within the subdivisions are also included and are found enlightening by the readers. Lists of Tables and Figures : Another device used to supplement the table of contents for throwing more light on the subject of the thesis is that of giving lists of tables and p~ures which occur in the report. 6.3.2 Main Body of the Report The main body of the research report contains all the material aspect of the research work. Introduction : The first part of the main body of the report, the Introduction, usually includes a statement of the factors leading up to the choice of the problem, the purposes of the study, the value and significance attached to the problem by the investigator as a contribution to knowledge and any other information to express the sincerity of the investigator in his selection. A statement and elucidation of the problem sometimes forms a part of the introduction; but more often/it is set up as a separate unit. If this is stated in a clear-cut and logical manner, the reader is able to get a sufficiently clear insight into the study from the very beginning. The problem should be defined in detail. The exact area the investigation is supposed to cover must be well demarcated. The sources of information selected and their nature and delimitations should be

mentioned and justified. All terms of a technical nature or those which may seem vague to the lay reader need to be defined carefully. The objectives, limitations, hypotheses, etc. are given. The methodology and design of the study are also given in introduction. To explain the developmental process used for the sturdy the investigator has to describe the techniques and tools he has used for collecting, organizing, analyzing and interpreting his data. The sources of data tapped, the channels prepared or adapted and utilized, the nature of data collected, their validity and reliability - all these should be given in a clear and adequate manner. Data collected, but rejected and the methods tried but not pursued - these should also find their place in the report and should not be just left out of the picture. Survey of Related Literature : Any research worker has to be up-todate in his information about studies, related to his own problem, already made by others. References are made to such similar or related studies and their evaluation too is made for the benefit of the reader either in the Introductory chapter, or else in a separate chapter. Herein the author finds another opportunity to justify his own endeavour and to emphasize the worthwhile elements in the treatment, selected by him, of the problem. Analysis and Interpretation ;The analysis and interpretation section deals with the main works undertaken. Each objective of the research work, each hypothesis, each research question posed and such other major constituents of the research work are thoroughly probed, analysed using the statistical data collected applying appropriate tools of analysis and interpretations are made in the light of the analysis made. Unusual or complex techniques of collection, organization, analysis and interpretation are explained in full. Whether the original data themselves should be included in the text or given in the appendix depends on the nature of the data. If they are not too extensive and are necessary to clarify the discussion, they should

certainly find a place in the text proper, or in the footnotes. If they are extensive and cumbersome, they should be placed in the appendix. Of the various aids used to make the presentation of data more effective, tables and figures are most common. When statistical data are assembled according to certain common factors in the form |of tables, significant relationships show up clearly. Depending on the type of material at one's disposal, many kinds of figures are found useful, e.g., statistical diagrams, photographs and maps, etc. All the information described above is sometimes confined to one chapter with separate subdivisions arranged stage-wise. Otherwise, separate chapters are devoted each major functional area or objective studied. The arrangement depends on the quantity of information one has to convey to the reader regarding the different stages in the process of the development of the study. Conclusion : The final unit of the report usually contains the findings of the study, the conclusions the investigator has arrived at, and the generalization he has formulated on the basis of the study. In stating the conclusions, the investigator must indicate what his contribution has been to his field of study. He should indicate on what data his various conclusions are based. He should clearly demarcate between the inevitable conclusions and his own interpretation of certain data. The range of applicability of the conclusions should be indicated on the basis of the limitations of the sources, the sample, the tools of collection and analysis, etc. Negative as well as positive results should find a place in the conclusions. Any recommendations, as to the application of the findings, the investigator wishes to make, can find a place in this chapter. Recommendations or suggestions for further study in the field touched by the present research are also found useful and are usually included in the concluding chapter.

6.3.3

Referencing Section of the Report Referencing section has three elements namely, bibliography,

appendix and index. Bibliography : The 'works cited' form of bibliography is preferable over the 'sources consulted'. Every book, thesis, article, documents which has been cited should be included in the list of 'works cited'. The bibliography should follow a logical arrangement in alphabetical order. In report of current practice is to have one comprehensive listing-not to divide into books, journals, newspapers, official papers, documents and manuscripts. The author(s) name, the title of the work, date of publication, name of the publisher and the place of publication be mentioned. For articles, the volume number and inclusive pages be also given, , The author's initials or surname should follow the name. When there are three or more authors of a particular work, the coauthors may be referred to as et.al, meaning and others. If there by more than one work by the same author, the author's name should be listed only once; subsequently a line will substitute his name. This bibliographical listing should not be numbered. It should be given only at the end of the thesis, Appendix : The appendix section gives a copy of the tools of research used, certain sample statistical workings, articles published by the researcher, etc. Each class of material given may be numbered as Appendix I, Appendix II and so on. It is saner to give the appendices in the same order in which the relevant items are used. Index : Index is a very important component which facilitates easy location of a concept or entity mentioned in the main body of the work. Here alphabetical order is followed. Page number is given to easy location. Author Index, Subject Index and Sponsor Index are

certain indices used. All the three may be separately given and merged into one single class of 'index1. 6.4 STEPS INVOLVED IN DRAFTING A RESEARCH REPORT A report must be well drafted so that it is seriously taken by others and all that have to be said are well said. Reader orientation, purpose orientation, time, orientation, technology orientation, etc are all needed. The steps in writing a research report are presented below. 6.4.1 Organization of Thought Organization of thought as to how the report be presented is the most fundamental starting point in the journey of preparation A research report. Ideas come before the mind eye. These are thought over again and a frame of presentation is planned. This plan does not in itself constitute style, but it is the foundation. 6.4.2 Acquaintance with the Research Full acquaintance with research is needed. This is facilitated by notes. It is desirable to make notes on separate cards or slips called form-facet. Accuracy and Adequacy are required - The second facet of mastery over notes consists in the investigator's complete control over the data, called study facet, i.e., understanding each fact by itself and in terms of others and of the researcher's own thoughts; the notes to be compared; criticized and revalued in order to enable the investigator to direct and organize the data in his own way and perhaps differently from what others have done.

6.4.3

First Draft The first draft concentrates on substance, i.e., fullness of facts.

All the facts of value are to be brought together. In addition to fullness, accuracy of the facts incorporated into the text becomes necessary. Another requirement is that there should be balance, proportion and development in facts. Importance is to be given to the comprehensiveness of the report but not to the language and form. For writing the -first draft the researcher should have control over his notes and should think continuously over the problem. In a way, the first draft is the most important of the different stages in reporting. This report may have to be re-written a number of times and still it continues to remain only a working draft. There are three purposes in writing the first draft, viz., to weave the material together for making clear connections, to assure the investigator himself of a satisfactory organizations and fullness of the facts, and to avoid blank paper fright that may be present in every young researcher. Second Draft: After a lapse of some time from the completion of first draft, the revision is made for writing the second draft. While drafting the second one, the researcher should concentrate largely on form and language. The researcher should give the first draft, at this stage, a shape so that it can be readable, clear and lucid. Considerable trimming or editing will have to be done to make the writing precise, concise and brief. Finally, at the second draft stage, critical evaluation will have to be made of all that has been written-facts, findings, conclusions and recommendations. To make the report readable and effective, the language plays major part. Third Draft : The final stage in drafting is the preparation of final report. It concentrates mainly on the finish and final touches, i.e. on documentation and polish to make the report weighty, authoritative, convincing and attractive. Documentation indicates the references to

the sources, other previous and current work and view, additional data and discussion and suggested further reading on the specific problem as handled by the researcher. In other words, it indicates the thoroughness of the investigation and on the other a guide to further work. A good research paper depends not only upon the amount of reading or notes taken or upon the form of presentation but also upon the accurate and thorough recording of the investigation. 6.5 PRINCIPLES OF GOOD REPORT WRITING Following are some important principles for writing a good research report: i) Make small sentences: Reading begins to get strenuous when sentences used in the research report average more than 25 words. ii) Vary sentence length: In using short sentences do not let the work become choppy. Sentences of considerable length are all right provided. Better they are balanced with enough short sentences. iii) Use simple words : The researcher is advised to use simple words in his research report. iv) Use familiar words: It is better to use familiar words in a research report. v) Avoid unnecessary words: The use of unnecessary words tire a reader and fog up the writing.

vi)

Write to express not to impress: The best way to impress the reader of report is to express what you have to say clearly and directly.

vii)

Write as you talk: The researcher should make his report writing as though it is his speech.

viii)

Keep as many active verbs as possible: Use of active verbs puts life into report writing.

ix)

Tie in with reader's experience: Always write research reports with a particular reader in mind. Relate what you have to tell him about vour research report. This is the way to have the reader understand your report.

x)

Make the report short and sweet : A short report makes reading interest and sweet. Short report should not mean short-cut report.

6.6

REFERENCING References must be explicitly given in the report. Bibliography is

one type of referencing and this has been already dealt in this lesson. . Other type of referencing is the foot notes and the same is presented below. Footnotes are meant to give complete bibliographical references and to provide the reader with information to enable him consult sources independently. These may be placed at the foot of a page or at the end of every chapter. When footnotes are given at the foot of the page, they are to be separated from the text by a fifteen space solid line drawn from the left margin and one double space below the last line of the text. If given at the end of a chapter, a centered heading

'FOOTNOTES' is necessary. Reference to footnotes is made by the use of superscripts i.e., numerals raised by one-half space.-i Footnotes should be numbered consecutively through a chapter, right after the statement. However, footnotes should not be resorted to common place statements or ordinary facts. They must be restricted to direct quotations, original ideas, statements, definitions, illustrations and diagrams. If a quotation found in one source that is obtained from another source, is used, both the sources should be listed in the footnote. Footnotes should be given in single line spacing with the first line only indented. A double space -should separate, successive footnotes. In this context, certain conventions in footnoting should be kept in mind. While writing the author's name, the first name or initials should be written first. Titles of completed works, such as books and reports should be underlined. Titles of articles should be enclosed by double quotation marks. The publisher's name, year or date of publication and the exact page(s) of the source of reference should also be given in this order. The meaning of accepted abbreviations meant to avoid repetitive arid lengthy documentation should be clearly understood, and should be used appropriately. Often employed abbreviations while footnoting include: ibid, loc.cit, op.cit., infra, supra, cf., and passim. 'Ibid' is the abbreviation of the word 'ibidem' which means, in the same place or work. This is to be used when two or more successive footnotes refer to the same source. However, if reference is to different pages, page numbers are to be indicated. When reference is made to the same place, that is the same page, as a preceding but not immediately preceding reference, loc.cit. is used following the last name of the author only. The abbreviation loc.cit. is derived from "loco citato" which means in the place cited. The abbreviated form of "opera

citato", viz. op.cit. should follow the author's name and precede the page reference of the same work as a preceding but not immediately preceding reference. To refer to the work already cited "supra" should follow the author's name and precede the page reference. 'Supra' means, 'above'. To refer to the text following 'infra' is to be used. To refer to the appendices or other annexure, the abbreviation Vide' which means 'see* is to be used. To invite the reader to compare what is stated with something else, the abbreviation cf. is to be used. When reference is to be made to opinions, attitudes etc., on a number of pages 'passim' - which means 'here and there' is to be used. The meaning of the above usages would be clear only if a reference is made to a standard publication. 6.7 CRITERIA FOR EVALUATING RESEARCH REPORT/

FINDINGS Aspects of the Report Criteria Needed i) ii) Title Problem Clarity and conciseness, suggest the problem and its scope clearly indicated a) b) c) d) e) f) g) h) i) j) iii) Methodology a) b) c) d) Worthy of researcher's effort Clearly stated Suitably sub-classified Suitably delimited Research objectives stated Research questions stated Research hypotheses stated Assumptions stated Terms defined Review of related works done Clear methodology Innovative methodology Improved methodology Validity of methodology established

iv)

Data Collection

a) b) c) d) e) f) g) h)

Data requirement clearly stated Adequate data collected Appropriate method used Population defined Sampling is scientific Tool is validated Proper administration of tool followed Checks and counter - checks effected Date are properly edited Data are classified well Data are tabulated well Data are presented in proper charts/ graphs Proper Statistical measures computed Hypotheses testing is appropriate Validity of statistical analyses established Lucid language followed Proper typology adopted Good grammatical approach followed Heading, footnotes, etc. are appropriate Punctuations, spelling, etc. are fine. Covers all the issues of the study Covers all the objectives Evaluations all the hypothesis Answers all the questions raised Neatly cauterized Paragraph and numbered are used Tables, Charts etc. are given appropriately at right places

v)

Data analysis

a) b) c) d) e) f) g)

vi)

Form and Style

a) b) c) d) e)

vii)

Main Body

a) b) c) d) e) f) g) h) i)

Foot notes are given in the right place Findings are appropriate and backed

up by supporting information j) k) l) viii) Findings a) b) Suggestions are rationale and practical Recommendations are valid Objectivity and unbiased ness followed through out. Findings must be beyond the obvious Findings must be in the light of the problems, c) d) e) objectives, research questions and hypotheses Findings must be backed by suitable reasoning and information. Findings must be based on appropriate statistical analysis and testing or logic. Findings must help solve problems of the society by at least pertinently defining the problem.

QUESTIONS : 1. 2. 3. 4. 5. 6. Explain the role and of research reports. Present the contents of a good research report Explain the steps involved in drafting research report. Present the principles of good reporting. What is referencing? Explain its need forms. What are the criteria for evaluating a research report and findings ?

REFERENCES 1. 2. 3. 4. 5. 6. 7. Research Material P. Saravanavel

Elements of Education Research - S.P.Sukhia Research for Marketing Decision - Green & Tull Marketing Research - Sangeetha Agarwal Research Methodology - Basetia and Sharma Encyclopedia of Modern Education Research - S.K. Sharma Research Methodology - C.R. Kothari

********

Model Question Paper Paper 2.1: RESEARCH METHODS

Time: 3 hours Maximum : 100 marks

PART-A Answer any Five questions 1. 2. 3. 4. 5. 6. 7. 8. Explain the process of research. Bring out the role of reviews of literature, How is a research design evalued?

(5x8 = 40)

Are probability sampling superior to non-probability sampling? How? Explain the factors affecting sample size. Assess interview as a method of data collection. Explain Likert Scaling technique. What is referencing? State its purpose. PART-B Answer any Four questions Question No. 15 is compulsory
(4x15 = 60)

9.

Develop a research design for a study on (i) consumer behaviour with reference to a particular class of product, or (ii) effectiveness of leadership style,

10. 11. 12. 13. 14. 15.

Explain the steps involved in writing a research report. What is interpretation? Give guidelines for valid interpretation. What are the types of questionnaire? Explain the need for types and sources of hypotheses. Explain the concepts of sampling and non-sampling error. Examine the uses of sampling error in research. A company considers changing the package colour and shape for its product. You are requested to conduct a market survey to help the company in zeroing in on possible range of colours and shapes for its product package. Explain how will you go about this work.