a) An element is that unit about which information is collected and that provides the basis of analysis. Elements are people or certain type of people. Other kinds of units can constitute the element for social research; families, social clubs. b) A population is the theoretically specified aggregation of study element. Where as the vague term. Eg all students of this university form population of the study of university crisis. c) A study population is that aggregation of elements from which the sample is actually selected. d) A sampling unit is that element or set of elements considered for selection in some stage of sampling. E.g. in a simple single stage sample the sampling units as the same as the elements. !or e"ample; you might select a sample of census blocks in a city then select a sample of house holds on the selected blocks and finally select a sample of adults from the selected house. #he sampling unit for three stages of sampling are respectively census, blocks, households and adult. e) A sampling frame is the actual list of sampling units from which the sample or some stage of the sample is selected. $f a simple sample of students is selected from a student roster is the sampling frame. f) An observation unit or unit of data collection is an element or aggregation of elements from which information is collected. g) %ariable is a set of mutually, e"clusively attributes se", age, employment, status and so&forth. #he element of a given population may be described in terms of their individual attributes on a given variable. A variable by definition must posses variation if all elements in the population have the same attribute that attribute is constant in the population rather than part of variable. h) A parameter is the summary description of given variable in a population. #he mean income of all families in a city and the age distribution of the city's population are parameters. i) A statistic is a summary description of a given variable in a sample. #hus the mean income computed from a sample and age distribution of that sample are statistics.
1. SAMPLE: (ample is a group of elements selected from a population in order to study that group instead of studying the whole population. Or is a group representing a population. E.g. if you want to study the crisis of university from the population of first year student, use some of them as sample. )
Why do we do stdy a sample! We study sample because economic savings in terms of money, time and resources e.g. it will not be possible to study all *+,(# students; you have to choose few of them according to the year, class, faculty or department. (o you will study the sample. We study sample and make analysis of the findings from the sample in order to know something about the entire population in which the sample is picked. We e"pect that what we find out about the sample will be true for the entire population because the characteristics of each member of the population are likely to be represented in the sample. $n other words sample studies are indicates of what e"ist in the population. $f the sample findings correspond to the population then that sample is representative of population. $f the sample findings don't correspond then the sample is known as unpresentative.
". SAMPLING #$E%&': (ampling theory is the study of relationships which e"ist between population and a sample drawn from that population. $n estimating unknown population parameters you use sample statistic to estimate population parameters. When you have a sample you can calculate sample statistics such as mean, median, mode, standardi-ation, validity etc. We select sample from the population and we use sampling frame which is the list of all elements in the population. We study sample because they have advantages a) #hey save time, money, effort. b) #hey are more convenient c) #hey help to get the information needed instead of using the whole population.
(. P&%)LEMS %* SAMPLING: #he findings from the sample may not be as valid as studying the whole population. #o do a sample survey the researcher must be involved in planning administration and programming survey however they involve in sample studies because sample present target population.
P&IN+IPLE %* S#,-'ING SAMPLING . ). +nity of the sample must be chosen in ob/ective and systematic manner. .. (ample units must be well defined to make it possible to identify them. 0. #he process of selecting sampling must be based on sound criteria and must avoid bias, distortion, and errors. 1. #he sample units should be independent of each other uniform and other same si-e and appear once in target population.
#'PES %* SAMPLING #here are two types of sampling A. *andom sampling 2probability sampling) 3. 4on&random sampling 24on&probability sampling) A- RANDOM SAMPLING
*andom sampling is done in a process in which every element in a determined population is in e5ual chances of being selected into the sample. #he selecting criteria are computed so that the selected sample represents the whole population. (election of such sample can be e"pensive, time consuming, complicated and it needs a large population in order to select your sample.
B- NON-RANDOM SAMPLING
4on&random sampling is a procedure which doesn't give e5ual chance to every element in the population to be selected into sample. #he sample selected is non&representative of the population. $t is the researcher who decides what elements to include in the sample. 0
A. #'PES %* P&%)A)ILI#' SAMPLING )) (imple random sampling .) (ystematic sampling. 0) (tratified random sampling. 1) ,luster sampling. 6) 7ultistage sampling. 8) Area sampling. 9) 7ultiphase sampling. :) ;anel studies sampling. <) (patial sampling.
1/ Simple random sampling: #his sampling techni5ue gives e5ual chance to every element of the population to be included in the sample. (election can be done by means of lottery method or random numbers or by means of computer method. a/ Lottery method: this involves identifying the sampling frame or list of all elements $n the target population and these elements are numbered or named in separate pieces of paper then they are mi"ed then thoroughly in a bo" or bucket then piece of paper is picked one by one fro the bo". And the picked numbers or names are recorded. #he picked number can be put into the basket or can be drawn again and picking continues until the re5uired number matched. 0/ &andom nm0ers: this method is similar to lottery method. #he difference is instead of using piece of paper you use random number which are found in the booklet. c/ +ompter method: this method is similar to the method of random numbers . the difference is instead of a researcher picking using the table of random numbers the researcher instructs the computer to pick those numbers.
"/ systematic sampling: $n this method selection of the sample unit is dependent o the selection of the previous one. !rom the target population the sample is selected in a systematic way by using sampling fraction.
Abbreviations target population = 4 sample si-e = n sampling fraction = k
#o select a sample you number from )& <. $f the population is students it means < students will be represented by one student. 2;icking of that student can be random or by chance provided you choose any student within < students ) 2suppose you pick student no 6) then you need a sampling fram<e from )&61.??? and you want a sample of 8.???. picking will be ),.,0,1,5,8,9,:,<,)?,)),).,)0,14,)6,)8,)9,):,)<,.?,.),..,,23,.1,.6,.8
Bour first picking in 6, second picking 6 C<=)1, third picking )1C<= .0 continously until you get 8,??? . sample out of 61.??? frame 2target population)
(/ strati1ied sampling: $n this method the researcher divide the population into strata e.g male@female, age groups, marital status etc. you can pick them by using either simple random or systematic random sampling. #he researcher divides a target population into number of strata according to number of the specific group of population to be included in the sample. #he researcher prepares a sampling frame for each data identified DstratumE from each strata and a sub& sample is drawn using one of the methods of random sampling. #he sub&sample can be proportional or disproportional. !or e"ample; if the population is )??? may be divided into three stratum. $) 7uslims 1?? ii) ,hristians 6?? iii) Findus )?? ;roportionally you may take )?G for each
7uslims 1? ,hristians 6? these are sub&samples Findus )?HHHHHHH )?? is sample
2/ +lster sampling: $n this sampling techni5ue you select cluster or groups of elements such as classes, schools and then each cluster you select a sub&sample using simple random sampling or systematic random sampling. Bou use cluster sampling when you do not have sampling frame for the whole population. Bou also use cluster sampling when the cluster criteria are important for the study cluster are economical. 6 #he problem of cluster sampling sample is based because they are picked from speciali-ed clusters. #he clusters may not be representative in the population. !or e"ample; a study of the university students of *+,(#. #wo clusters chosen are (A(( and ($E. !rom each cluster you select a sample randomly. #he two sub&samples are combined to form a sample for the study. #hat is because it omits other faculties in *+,(#.
3/ Mlti.stage sampling: 7ulti stage sampling method involves drawing samples. #o do this you prepare a sampling of the target population and then draw a large sample using simple random sampling or systematic random sampling and you get the primary sample. Bou continue to draw another sample until you get the re5uired sample.
(tage $ = ).??? from )?.??? (tage $$ = 6?? from ).??? (tage $$$ = )?? from 6??
(o you may pick until you get a small si-e. #his can be done through interviews. Bou may use simple random or systematic random sampling. Advantage of this procedure is that; as you pick one sample form the previous sample, you make the sample more specific and more relevant to the research 5uestions. #he sample becomes more representative of the target population
4/ Area sampling: #his sampling method is like a multi&stage sampling applied in a geographical area. E.g. the study of A$I( or F$% in the world. $ & Africa J from Africa you choose $$ & #an-ania J form #an-ania you choose $$$& Iar es salaam
#he sample of the study is the last stage. (ample which is drawn in a series of samples taken from geographical areas and each stage refers progressively. Area sampling is a form of cluster sampling because procedures give high representation from as many geographical areas as possible.
5/ Mlti.phase sampling: #his is like multi&stage sampling in that primary selection is done and from that another selection is done, e.g. target population )?.??? students ;hase $ = study )??? students ;hase $$ = study 6?? students ;hase $$$ = study )?? students 8
Bou can use simple or systematic sampling each phase, conduct a study and complete before you move to another phase. Advantage of this type of sampling is that; the information found in earlier phase will help to select the relevant and representative sample.
6/ Panel stdies: #hese are longitudinal studies in which the respondents of the original sample are studied more than one occasion. E.g. studying coffee farmers in Kuaso district. Bou go to study in every 6 years in the same population. Like ) st study )?.??? farmers )<<< . nd study )?.??? farmers .??1 0 rd study )?.??? farmers .??<
Pro0lem o1 condcting longitdinal: #he problem which arises to this & 3ecause your ob/ective is to find out if there is continuity of over time the problem is whether or not. Bou will be able to get all people in different times. & #he problem of mortality, migration and changing of residence. & As they get used to the study they can give wrong information deliberately. & #he problem of motivation.
7/ Spatial sampling: #his method re5uires the target population to assemble as a crowd in a space. #he interviewer approaches to the crowd and chooses the respondents. E.g. you go to the crowd of !aculty of Arts. Bou move maybe systematically two steps. Bou choose a respondents repeated or far steps after four steps you get respondent until you get a sample si-e.
). #'PES %* N%N.P&%)A)ILI#' SAMPLING 4on&probability sampling procedures don't produce samples which presents the target population. 4on&probability sampling is mostly used in 5ualitative research and in research in which representativeness is not considered in a study. #he most common types of the non&probability sampling are; ). Accidental sampling .. ;urposive sampling 0. Muota sampling 1. (nowball sampling 9
1/ Accidental sampling: #his is also known as incidental sampling, chunk sampling, and hapha-ard sampling. $n this sampling techni5ue the researcher interviews all the respondents for that study who comes in contact accidentally with during the research time.
"/ Prposi8e sampling: $n this sampling techni5ue the researcher chooses respondents who in their opinion are thought to be relevant to the sub/ect under the study. #o do this the researcher must identify the respondent and arrange the time and place to meet them and interview. Bou have to identify people who know the problem like; dean of students or head of Iepartment or chairperson etc. you have to negotiate with them in order to get the proper time and place where to meet.
(/ 9ota sampling: #his is a kind of stratified sampling. #he differences is that instead of dividing the population into strata and choose the respondent randomly, you set up 5uotas of respondents to be chosen from specific population groups. !or e"ample; primary schools' pupils in /oining ((( ) have to be divided into 5uotas like; Iansoman = .6? Odorkor = 8?? Lapa- = ).??? Achimota = 1??
Bou define your basis of choices according to gender, education, or marital status. Bou set up 5uotas to each group. $n setting the 5uota, will depend on the nature and the type of the research. #he choice is left to the interview. #he researcher must decide how to set the 5uota which can be proportional or disproportional.
2/ Snow0all sampling: #he researcher begins with a few respondents who are available. After interviewing they ask the respondent to recommend other person who meets the criteria of the research and who are willing to participate in the study. Bou continue until you reach the re5uired sample. Advantages and disadvantages of case studies Advantages : 1 Case study data are drawn from peoples experiences and practices and so are seen to be strong in reality. 2 Case studies allow for generalizations from a specific instance to a more general issue. 3 Case studies allow the researcher to show the complexity of social life. Good case studies build on this to explore alternative meanings and interpretations. Case studies can provide a data source from which further analysis can be made. !hey can" therefore" be archived for further research wor#. $ %ecause case studies build on actual practices and experiences" they can be lin#ed to action and their insights contribute to changing practice. &ndeed" case study may be a subset of a broader action research pro'ect. ( %ecause the data contained in case studies are close to peoples experiences" they can be more persuasive and more accessible. Source) *dapted from Cohen and +anion ,1--$) 123.. Disadvantages !he disadvantages of case studies are lin#ed to their advantages. &n particular) 1. !he very complexity of a case can ma#e analysis difficult. !his is particularly so because the holistic nature of case study means that the researcher is often very aware of the connections between various events" variables and outcomes. *ccordingly" everything appears relevant. /ou need to show the connections but not lose sight of the whole. 2. 0hile the contextualization of aspects of the case strengthen this form of research" it is difficult to #now where 1context begins and ends. One other caution about adopting a case study approach to your research pro/ect has already been mentioned in this book. $t also applies to action research, as well as to other research approaches. #his concerns the tendency for some students, particularly those in employment to base their research pro/ects within their places of employment. Experiments < E"periment is a situation in which the independent variable 2also known as the e"posure, the intervention, the e"perimental or predictor variable) is carefully manipulated by the investigator under known, tightly defined and controlled conditions, or by natural occurrence. At its most basic, the e"periment consists of an experimental group which is e"posed to the intervention under investigation and a control group which is not e"posed. #he e"perimental and control groups should be e5uivalent, and investigated systematically under conditions that are identical 2apart from the e"posure of the e"perimental group), in order to minimi-e variation between them. 23owling )<<9 )<)) !or many people undertaking small&scale research in the social sciences, the idea of conducting e"periments may seem rather strange. #he e"perimental method is particularly associated with the physical sciences, where materials and nonhuman life forms are more amenable to e"perimentation. $ndeed, e"periments are at the heart of what is known as the scientific method, with its practice of formulating and testing hypotheses through carefully designed and controlled tests. #he associated terminology, repeated in the 5uotations above J e.g. sub/ects, independent, dependent and e"traneous variables, and control J appears very precise and suggestive. E"periments are, however, widely used as a research approach in a number of the social sciences, particularly psychology 2which is often classified as a science Advantages and disadvantages of experiments Advantages 1 %ecause of the random assignment of people to intervention and control ,i.e. randomization of extraneous variables. the ris# of extraneous variables confounding the results is minimized. 2 Control over the introduction and variation of the 1predictor variables clarifies the direction of cause and effect. 3 &f both pre2 and post2testing is conducted this controls for time2 related threats to validity. !he modern design of experiments permits greater flexibility" efficiency and powerful statistical manipulation. $ !he experiment is the only research design which can" in principle" yield causal relationships. Disadvantages 1 &t is difficult to design experiments so as to represent a specified population. 2 &t is often difficult to choose the 1control variables so as to exclude all confounding variables. )? 3 0ith a large number of uncontrolled" extraneous variables it is impossible to isolate the one variable that is hypothesized as the cause of the other3 hence the possibility always exists of alternative explanations. Contriving the desired 1natural setting in experiments is often not possible. $ !he experiment is an unnatural social situation with a differentiation of roles. !he participants role involves obedience to the experimenter ,an unusual role.. ( 4xperiments cannot capture the diversity of goals" ob'ectives and service inputs which may contribute to . . . outcomes in natural settings. ,%owling 1--5) 1-3.. Surveys Survey research . . . is the method of collecting information by asking a set of pre formulated questions in a predetermined se5uence in a structured questionnaire to a sample of individuals drawn so as to be representative of a defined population. 2Futton )<<? :) (urvey research involves the collection of information from members of a group; e.g.; students, teachers, or other persons associated with an issue in society and the analysis of this information to illuminate important societal issues. 7ost surveys are based on samples of a specified target population J the group of persons in whom interest is e"pressed. #he researcher often wishes to generali-e the results obtained from the samples to the populations from which the samples were drawn 2*osier )<:: )?9). As these 5uotations indicate, surveys are usually associated as a research approach with the idea of asking groups of people 5uestions. Advantages and disadvantages of surveys Advantages 6 7uestions are designed so that answers from individual interviews can beadded together to produce results which apply to the whole sample . . . 6 !he research is based on interviews with a representative sample of respondents. . . 6 !he 8uestions are designed to be unbiased . . . 6 9urveys lend themselves to future replication . . . 6 :arge surveys can often be bro#en down. Disadvantages 1 !he data" in the form of tables" pie charts and statistics" become the main focus )) of the research report" with a loss of lin#age to wider theories and issues. 2 !he data provide snapshots of points in time rather than a focus on the underlying processes and changes. 3 !he researcher is often not in a position to chec# first hand the understandings of the respondents to the 8uestions as#ed. &ssues of truthfulness and accuracy are thereby raised. !he survey relies on breadth rather than depth for its validity. !his is a crucial issue for small2scale researchers. #en reasons for reading for research ) 3ecause it will give you ideas. . 3ecause you need to understand what other researchers have done in your area. 0 #o broaden your perspectives and set your work in conte"t. 1 3ecause direct personal e"perience can never be enough. 6 #o legitimate your arguments. 8 3ecause it may cause you to change your mind. 9 3ecause writers 2and you will be one) need readers. : (o that you can effectively critici-e what others have done. < #o learn more about research methods and their application in practice. )? $n order to spot areas which have not been researched. &ESEA&+$ :A&IA)LES !he #ey to designing any experiment is to loo# at what research variables could affect the outcome. * researcher must then determine which variable needs to be manipulated to generate 8uantifiable results. !here are many types of variable but the most important" for the vast ma'ority of research methods" are the independent and dependent variables. %ery simply, a :A&IA)LE is a measurable characteristic that varies 2i.e.; something that changes). $t may change from group to group, person to person, or even within one person over time. * variable changes according to different factors. ). (ome variables change easily, like the stock&e"change value, while other variables are almost constant, like the name of someone. A variable can be a number, a name or anything where the value can change. *n example of a variable is temperature. !he temperature varies according to other variables and factors. /ou can measure different temperatures inside and outside. &f it is a sunny day" chances are that the temperature will be higher than if it;s cloudy. *nother thing that can ma#e the temperature change is whether something has been done to manipulate the temperature" li#e lighting a fire in the chimney. &n research" you typically define variables according to what you;re measuring. !he independent variable is the variable which the researcher would li#e to measure ,the cause." while the dependent variable is the effect ,or assumed effect." which is dependent on the independent variable. !hese variables are often stated in e"perimental research" in a hypothesis" e.g. <what is the effect of personality on helping behavior=< #B;E( O! %A*$A3LE( -EPEN-EN# :A&IA)LES #he dependent variable is the measurable outcome of manipulation, the results of the e"perimental design. Iependent variables show the effect of manipulating or introducing the independent variables. !or e"ample, if you designed an e"periment to determine how 5uickly a cup of coffee cools, the manipulated independent variable is time and the dependent measured variable is temperature. $n other words, the variation in the dependent variable depends on the variation in the independent variable. IN-EPEN-EN# :A&IA)LES #he independent variable is the core of the e"periment and is isolated and manipulated by the researcher. #hey are those that the researcher has control over. #his NcontrolN may involve manipulating e"isting variables 2e.g., modifying the temperature of a free-er) or introducing new variables 2e.g., adding a bottle of oil to water in the free-er) in the research setting. Whatever the case may be, the researcher e"pects that the independent variable2s) will have some effect on 2or relationship with) the dependent variables. )0 -A#A Another very important component is the -A#A. Iata come in various types. #hey are a representation of reality, and show the results of measuring properties or processes. Iata and the ways they are measured come in various types. N%MINAL -A#A 4ominal means Nname bearing.N #he nominal scale places things into named categories. #hese things are assigned to groups according to their common or shared elements. !or e"ample women who are different in many ways could be assigned to the same category based on their shared gender. $mportant #he different categories are not ordered in any Nmore or lessN sense. #hey are /ust different from each other. %&-INAL -A#A #he ordinal scale places things in order. Ordinal data show a particular itemOs position relative to other items, such as N!irst, second, third, etc.N #he ordinal scale doesnOt specify the distance between each item. $t /ust puts them in order. !or e"ample, in a playground foot race where no one has a watch, the participants will not know their actual times. #hey will only know who came in first, second, or third. IN#E&:AL -A#A #he interval scale uses e5ual&si-ed units of measurement 2points, minutes, etc.) and, therefore, shows the distances, or intervals, between sub/ectsO performances. $n the foot race e"ample, if the runnersO classmates started counting aloud after the first runner crossed the finish line, they might discover that the second place finisher was only two counts behind the winner while the third&place finisher was ten counts behind. $nterval data show this difference in distances. Ordinal data would not. $t is important to remember that with interval 2as opposed to ratio) data, the intervals start from an arbitrary point, not absolute -ero. #herefore, a student who scores a 8? on a grammar test could not be said to know twice as much grammar as a student who scored 0?. Also, the person who scored a ? on this test would not be said to know no grammar at all. &A#I% -A#A )1 #he ratio scale is like the interval scale. $t employs e5ual intervals. Fowever, the ratio scale begins at a true -ero point. #hat point represents an absolute lack of the 5uality being measured. 3ecause of this characteristic, additional mathematical functions are possible with ratio data that are not possible with other types of data. Note (trictly speaking, the word data is actually a plural. #he singular form 2which is rarely used) is datum. #herefore, it is correct to say, for instance, N#he data showN or N#he data areN rather than N#he data showsN or N#he data isN. S,);E+#S S,);E+#S are the sources of your data. !or instance most research in language&related fields uses people as sub/ects. #heir characteristics, development, opinions, attitudes, knowledge, performance, etc. are used to answer your research 5uestion. $n order to choose appropriate sub/ects you need to decide what your population of interest is. P%P,LA#I%NS $n research, population has a speciali-ed meaning. #heoretically, a population is the group from which your sub/ects are drawn. #herefore, it is also the group that your sub/ects represent. When discussing your research findings, you must be careful not to generali-e your conclusions beyond this group. !or this reason, it important to identify the key characteristics of your sub/ects, and the population they represent. Are they children, teenagers, or adultsP Are they men or womenP What is their nationalityP What native language do they speakP INS#&,MEN#S < P&%+E-,&ES INS#&,MEN#S are used to gauge some 5uality or ability of your sub/ects. #he purpose of the instrument is to elicit the data for your study. $ere are some possi0le instrments=procedres: $nterviews 2unstructured or structured) Muestionnaires 2mailed or in&person) Observations )6 Iiaries kept *eviews of records or documents %erbal self&reports by respondents ANAL'>ING -A#A Once have your data, you must ANAL'>E it. #here are many different ways to analy-e data some are simple and some are comple". (ome involve grouping, while others involve detailed statistical analysis. #he most important thing you do is to choose a method that is in harmony with the parameters you have set and with the kind of data you have collected. IN#E&P&E#A#I%N %* #$E &ES,L#S )8 >esearch does re8uire some interpretation and extrapolation of results. &n scientific research" there is always some #ind of connection between data ,information gathered. and why the scientist thin#s that the data loo#s as it does. ?ften the researcher loo#s at the data gathered" and then comes to a conclusion of why the data loo#s li#e it does. )9