Académique Documents
Professionnel Documents
Culture Documents
e-ISSN: 2319-2380, p-ISSN: 2319-2372. Volume 5, Issue 2 (Sep. - Oct. 2013), PP 51-58
www.iosrjournals.org
Abstract: Experimental research is important to society - it helps us to improve our everyday lives.Typically,
an experiment is constructed to be able to explain some kind of causation. This article is focused onvarious
issues involved in comparing related groups and measuring change with pretest and posttest data. Different
pretest-posttest designs are discussed in a manner that can help researchers to select suitable test in order to
maintain internal and external validity simultaneously. One real time example from agricultural sector is given
for better understanding of theme of the paper. Despite its complexity in analysis, Solomon four group methods
help the researchers in taking valid conclusions in any field of Research. Findings enumerated for each
methodcertainly provide guidance in drawing conclusions appropriately.
Keywords: Solomon four group test, pretest-posttest designs, internal &external validity
I.
Introduction
Experimental designs are often touted as the most "rigorous" of all research designs or, as the "gold
standard" against which all other designs are judged. True experimental design is regarded as the most accurate
form of experimental research, in that it tries to prove or disprove a hypothesis mathematically, with statistical
analysis[3].In a true experimental study, a control and an experimental group are used. The dependent variable
in each group is observed before introduction of the independent variable. Then only the experimental group is
introduced to the independent variable. Observations of the dependent variable are made for both groups after
the exposure of the experimental to the independent variable.
Pretest-posttest designs are an expansion of the posttest only design with nonequivalent groups, one of
the simplest methods of testing the effectiveness of a treatment. In this design, which uses two groups, one
group is given the treatment and the results are gathered at the end. The control group receives no treatment,
over the same period of time, but undergoes exactly the same tests. Statistical analysis can then determine if the
intervention had a significant effect or not. One common example of this is in agriculture; one set of plots is
given a fertilizer, whereas the control group is given none, and this allows the researchers to determine if the
fertilizer really works. This type of design, whilst commonly using two groups, can be slightly more complex.
For example, if different levels of a fertilizer are tested, the design can be based around multiple groups.
Whilst this posttest only design does find many uses, it is limited in scope and contains many threats
to validity. It is very poor at guarding against assignment bias, because the researcher knows nothing about the
individual differences within the control group and how they may have affected the outcome. Even with
randomization of the initial groups, this failure to address assignment bias means that the statistical power is
weak.The results of such a study will always be limited in scope and, resources permitting; most researchers use
a more robust design, of which pretest-posttest designs are one. The posttest only design with non-equivalent
groups is usually reserved for experiments performed after the fact, such as anAgricultural researcher wishing to
observe the effect of a fertilizer that has already been administered.
For many true experimental designs, pretest-posttest designs are the preferred method to compare
participant groups and measure the degree of change occurring as a result of treatments or interventions.Pretestposttest designs grew from the simpler posttest designs, and address some of the issues arising with assignment
bias and the allocation of participants to groups. One example is Agriculture, where researchers want to monitor
the effect of insecticides upon yields of wheat. Other areas include evaluating the effects of counseling, testing
medical treatments, and measuring psychological constructs. The only stipulation is that the subjects must
be randomly assigned to groups, in a true experimental design, to properly isolate and nullify any nuisance
or confounding variables.
www.iosrjournals.org
51 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
II.Terminology
2.1 Confounding Variable
In any experiment there are many kinds of variables
that will affect the experiment. The independent variable is the
manipulation for the experiment and the dependent variable is
the measurement taken from that experiment. Confounding
variables are things in which have an effect on the dependent
variable, but were taken into account in the experimental
design.
For example, one wants to know if fertilizer X has an effect on getting more yields. The experimenter
must take care to design the experiment so that he can be very sure that the yields under observation must be
gained because of the influence of Fertilizer X, and that the improvement was not caused by other factors, called
confounding variables.
2.2 Internal and External validity
External and internal validity are not all-or-none, black-and-white, present-or-absent dimensions of an
experimental design. Validity varies along a continuum from low to high.
2.2.1. Internal validity is a property of scientific studies which reflects the extent to which a causal conclusion
based on a study is warranted. Such warrant is constituted by the extent to which a study minimizes bias. In
other words, to the degree that we are successful in eliminating confounding variables within the study itself is
referred to as internal validity.
In other words, Internal validity is the degree to which the experimental treatment makes a difference
in (or causes change in) the specific experimental settings [5]. The factors that threaten internal validity are:
history, maturation, pretest effects, instruments, and statistical regression toward the meanand differential
selection of participants, mortality, and interactions of factors (e.g., selection and maturation).
2.2.2 External validity is the validity of generalized (causal) inferences in scientific studies, usually based on
experiments as experimental validity. In other words, it is the extent to which the results of a study can be
generalized to other situations and to other people. Hence, a study that readily allows its findings to generalize
to the population at large has high external validity [14].Hence, External validity is the degree to which the
treatment effect can be generalized across populations, treatment variables, and measurement instruments.
Threats to external validity include: interaction effects of selection biases and treatment, reactive interaction
effect of pretesting, reactive effect of experimental procedures, and multiple-treatment interference [1].
52 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
3.2 The Two Group Control Group Design
This is, by far, the simplest and most common of the pretest-posttest designs, and is a useful way of
ensuring that an experiment has a strong level of internal validity. The principle behind this design is relatively
simple, and involves randomly assigning subjects between two groups, a test group and a control. Both groups
are pre-tested, and both are post-tested, the ultimate difference being that one group was administered the
treatment.
This test allows a number of distinct analyses, giving researchers the tools to filter out experimental
noise and confounding variables. The internal validity of this design is strong, because the pretest ensures that
the groups are equivalent. The various analyses that can be performed upon a two-group control group pretestposttest designs are (Fig 1).
53 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
In the figure, A, A1, B and C are exactly the same as in the standard two group design.
IV. Methodology
4.1 Research approach
Survey method has been adopted to study the key objective of the Research i.e. whether training
programme for farmers conducted by Govt.Officials on modern cultivation methods is effective or not.
www.iosrjournals.org
54 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
4.2 Population
The population for the present study consists of large farmers who had more than 10 acres of cultivable
land in Chittoor district of Andhra Pradesh.
4.3 Sample and Sampling Technique
Three Stage Random sampling technique has been adopted to select the sample of farmers. For this
study four mandals have been selected and from each mandal one village has been selected randomly. From
each selected village, 24 farmers are again selected randomly. Hence, totally 96 farmers in four groups of size
24 haven been selected at random.
4.4 Study variable is Awareness score of farmers on modern cultivation.
A tool has been prepared to assess the knowledge levels of farmers with regard to modern cultivation
methods. Tool consists of 50 questions and measured with Yes or No type options. Reliability is found as
8.65 which is good and validity is checked.
4.5 Data
A study is designed to assess the awareness of farmers on modern cultivation techniques in agriculture.
As discussed in methodology, four groups of farmers had been considered and groups are labeled 1, 2, 3 and 4
respectively. Out of four groups, Group-2 and Group-4 have not been invited for awareness programme on
modern cultivation methods conducted by Govt. Officials. Group-1 and Group-3 had attended an awareness
programme. Group-1 and Group-2 had been interviewed before and after training whereas Group-3 and Group-4
had been questioned only after the training. Awareness scores are noted in table-1.
Notations:
Ex1Pre: Awareness Scores of group-1before attending training.
Ex1Post: Awareness Scores of group-1after attending training.
CG1Pre: Awareness Scores of group-2 before training who are not invited for training.
CG1Post: Awareness Scores of group-2 after training who are not invited/attended.
Ex2Post: Awareness Scores of group-3 after attending training.
CG2Post: Awareness Scores of group-4 after training who are not invited/attended.
Table-1: Awareness score of the farmers
Group-1
Group-2
Farmer
Number
Ex1Pre
Ex1Post
CG1Pre
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Mean
29
25
30
27
29
29
27
28
26
29
26
27
29
28
30
26
30
27
30
30
27
30
25
27
27.96
32
30
34
34
31
32
35
32
34
30
32
35
32
30
32
31
34
33
32
30
32
30
31
35
32.21
26
26
25
29
29
29
29
25
26
25
30
27
28
26
25
26
28
30
30
27
25
26
29
29
27.29
Group-3
Group-4
CG1Post
Ex2Post
CG2Post
27
27
29
27
27
29
28
30
27
27
30
29
27
30
26
26
28
26
27
27
28
30
28
26
27.75
35
35
32
31
32
31
34
35
33
35
34
35
35
32
35
35
34
35
34
33
33
34
33
34
33.71
30
25
28
27
25
30
26
29
27
28
25
30
29
26
25
25
25
30
27
29
29
27
27
25
27.25
www.iosrjournals.org
55 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
4.6 Statistical Analysis
Collected data is properly entered in computer for the sake of analysis. Appropriate Statistical
techniques such as Independent sample t-test and Paired t-tests have been applied with the help of SPSS version
20 and results are concluded as per the level of significance. Significant findings are presented graphically.
V.
Three different designs had been applied in order to test the main objective and findings have been
compared with respect to their effectiveness.
5.1
Solution with the Posttest Only Design with Non-Equivalent Control Groups
Line/Fig
Test No
D/Fig(2)
Group
CG2POST
EX2POST
Mean
27.25
33.71
Sd
1.89
1.33
Mean Diff
t-value
p-value
Remark
6.46
13.66**
0.00
Sig @ 1%
Test No
Group Mean
EX1PRE 27.96
A /Fig(1) 1(Pt-test)
EX1POST 32.21
CG1PRE 27.29
A1/Fig(1) 2(Pt-test)
CG1POST 27.75
CG1PRE 27.29
B/Fig(1) 3(It-test)
EX1PRE 27.96
CG1POST 27.75
C/Fig(1) 4(It-test)
EX1POST 32.21
*significant @ 5% level
** significant @ 1% level
It-test : Independent sample t-test
Pt-test : Paired/related sample t-test
Sd
1.68
1.69
1.83
1.36
1.83
1.68
1.36
1.69
Mean Diff
t-value
p-value
Remark
4.25
8.25**
0.00
Sig @ 1%
0.46
0.93
0.36
NS
0.67
1.31
0.20
NS
4.46
10.06**
0.00
Sig @ 1%
www.iosrjournals.org
56 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
5.3 Solution with the Solomon four Group Control Group Design
Line
Test No
A/Fig(2)
1(Pt-test)
A1/Fig(2)
2(Pt-test)
B/Fig(2)
3(It-test)
C/Fig(2)
4(It-test)
D/Fig(2)
5(It-test)
E/Fig(2)
6(It-test)
F/Fig(2)
7(It-test)
G/Fig(2)
8(It-test)
Group
EX1PRE
EX1POST
CG1PRE
CG1POST
CG1PRE
EX1PRE
CG1POST
EX1POST
CG2POST
EX2POST
CG1PRE
CG2POST
EX1POST
EX2POST
CG1POST
CG2POST
Mean
27.96
32.21
27.29
27.75
27.29
27.96
27.75
32.21
27.25
33.71
27.29
27.25
32.21
33.71
27.75
27.25
Sd
1.68
1.69
1.83
1.36
1.83
1.68
1.36
1.69
1.89
1.33
1.83
1.89
1.69
1.33
1.36
1.89
Mean Diff
t-value
p-value
Remark
4.25
8.25**
0.00
Sig @ 1%
0.46
0.93
0.36
NS
0.67
1.31
0.20
NS
4.46
10.06**
0.00
Sig @ 1%
6.46
13.66**
0.00
Sig @ 1%
0.04
0.08
0.94
NS
1.50
3.41**
0.00
Sig @ 1%
0.50
1.05
0.30
NS
34
32
28
EX1PRE
EX1POST
Group-1
27
28
27
CG1PRE
CG1POST
CG2POST
EX2POST
Group-3
Group-4
Group-2
www.iosrjournals.org
57 | Page
Significance of Solomon four group pretest-posttest method in True Experimental Research- A Study
VI. Conclusion
Experiments are conducted to be able to predict the phenomenon. Present scenario of the research is
mainly based on experimental research which provides causality for the variation. There are several designs to
find causal relationship among variables. Amongst, Solomon four-group design(Design-3.3) is superior to
Posttest only design(Design-3.1) and Two group control group designs(Design-3.2) because, along with
controlling for effects of history, maturation, and pretesting, it allows for evaluation of the magnitudes of such
effects and higher degree of external validity despite difficulty in conducting the design.
Acknowledgements
Author is very grateful to the farmers, and farm workers who assisted wholeheartedly in collecting the
data from them at Nellore district.
References
[1]
J. Bellini and P. Rumrill, Research in rehabilitation counseling, Springfield, IL: Charles C.Thomas.
[2]
R.D. Bock, Basic issues in the measurement of change. in: Advances in Psychological and Educational Measurement, D.N.M.
DeGruijter and L.J.Th.Van der Kamp, eds, JohnWiley&Sons, NY, 1976, pp. 7596.
[3]
A.D. Bryk and H. I. Weisberg, Use of the nonequivalent control group design when subjects are growing, PsychologicalBulletin85
(1977), 950962.
[4]
I.S. Cahen and R.L. Linn, Regions of significant criterion difference in aptitude- treatment interaction research,
AmericanEducational Research Journal 8 (1971), 521530.
[5]
L.J. Cronbach and L. Furby, How should we measure change - or should we? PsychologicalBulletin74 (1970), 6880.
[6]
D.M. Dimitrov, S. McGee and B. Howard, Changes in students science ability produced By multimedia learning environments:
Application of the Linear Logistic Model for Change, School Science and Mathematics 102(1) (2002), 1522.
[7]
G.H. Fischer, Some probabilistic models for measuring change, in: Advances in Psychological and Educational
Measurement,D.N.M. DeGruijter and L.J.Th. Van der Kamp, eds,John Wiley & Sons, NY, 1976, pp. 97110.
[8]
G.H. Fischer and E. Ponocny-Seliger, Structural Rasch modeling, Handbook of the usage of LPCM-WIN 1.0,
Progamma,Netherlands, 1998.
[9]
R.K. Hambleton, H. Swaminathan and H. J. Rogers, Fundamentals of Item Response Theory, Sage, Newbury Park, CA,1991.
[10]
S.W. Huck and R.A. McLean, Using a repeated measures ANOVA to analyze data from A pretest-posttest design: A potentially
confusing task, Psychological Bulletin 82 (1975),511518.
[11]
S. Isaac and W.B. Michael, Handbook in research and evaluation 2nd. ed., EdITS, San Diego, CA, 1981.
[12]
E. Jennings, Models for pretest-posttest data: repeated measures ANOVA revisited, Journal of Educational Statistics 13 (1988),
273280.
[13]
K.G. Joreskog and D.Sorbom, Statistical models and methods for test-retest situations, in: Advances in Psychological
andEducational Measurement, D.N.M. DeGruijter and L.J.Th.Van der Kamp, eds, John Wiley & Sons, NY, 1976, pp. 135157.
[14]
L. Linn and J.A. Slindle, The determination of the significance of change between pre- and posttesting periods, Review of
Educational Research 47 (1977), 121150.
[15]
F.M. Lord, The measurement of growth, Educational and PsychologicalMeasurement16 (1956), 421437.
[16]
S. Maxwell, H.D. Delaney and J. Manheimer, ANOVA of residuals andANCOVA: Correcting an illusion by using model
comparisons and graphs, Journal of Educational Statistics 95 (1985), 136147.
[17]
G.J. Mellenbergh, A note on simple gain score precision, AppliedPsychologicalMeasurement 23 (1999), 8789.
[18]
J.E. Overall and J. A. Woodward, Unreliability of difference scores: A paradox for measurement of change,
PsychologicalBulletin82 (1975), 8586.
[19]
D. Rogosa, D. Brandt and M. Zimowski, A growth curve approach to the measurement of change, Psychological Bulletin 92 (1982),
726748.
[20]
I. Rop, The application of a linear logistic model describing the effects of preschool education on cognitive growth, in: Some D.M.
Dimitrov and P.D. Rumrill, Jr. / Pretest-posttest designs and measurement of change 165 mathematical models for social
psychology, W.H. KempfandB.H. Repp, eds, Huber, Bern, 1976.
[21]
D. Sorbom, A statistical model for the measurement of change in true scores, in: Advances in Psychological and Educational
Measurement, D.N.M. DeGruijter and L.J.Th. Van der Kamp, eds, John Wiley & Sons, NY, 1976, pp. 11591170.
www.iosrjournals.org
58 | Page