I was in the midst of reading Creswell’s chapter 10 (experimental designs) when I noticed that he does not describe the Solomon four-group design. In my library of 17 research textbooks, only two describe this between-groups design. Additionally, information on the Internet is sparse, and it does not have a Wikipedia page.
The Solomon four-group design can be used to mitigate threats to external validity, particularly the affects pretesting has on the overall study. This may also lessen the Hawthrone effect, where participants are motivated to improve because they are part of a study.
To be valid, this design must have randomly assigned participants. According to Trochim and Donnelly (2008) two groups receive the treatment, or stimulus, and two do not. Additionally, two groups are pretested and two are not (p. 203). This is more easily understood visually.
For Group 1, the affect of the treatment can be measured from the pretest to the posttest. In Group 2, also known as the control group, there should be little or no change. Thus, we can compare the affect of the treatment between groups 1 and 2 by means of the results of the posttests.
Groups 3 and 4 are examined next, with Group 3 receiving the treatment and Group 4, the control group, not receiving the treatment. The posttests from these two groups are then compared for affect.
If we now turn our attention to Creswell (2012), he has very nearly described the Solomon four-group design in table 10.3 on page 310. If he had combined the pre- and posttest design with the posttest-only design we could easily see the Solomon four-group design.
According to Babbie and Benaquisto (2002) the findings from the posttests of groups 3 and 4 “rule out any interactions between the testing and the stimulus” (p. 227). This is true for all random assigned posttest-only designs. Babbie and Benaquisto go on to explain that a benefit to this design is that
it also provides data for comparisons that will reveal the amount of such interaction that occurs in the classical experimental design. This knowledge allows a researcher to review and evaluate the value of any prior research that used the simpler design (p. 227).
Changing Minds (2013) states that “for reliable results, several sets of four tests should be applied and the means measured.” However, we must remember that this design takes more time than “regular” true experimental designs because it takes more time, presumably more money, and more participants. This is particularly true if more than one Solomon four-group is employed, as suggested by Changing Minds.
Babbie, E., & Benaquisto, L. (2002). Fundamentals of social research, first Canadian edition. Toronto, Canada: Thomson Nelson.
Changing Minds. (2003). Solomon four-group design. Retrieved from http://changingminds.org/explanations/research/design/solomon.htm
Creswell, J. W. (2012). Educational research: Planning, conducting, and evaluating quantitative and qualitative research. Boston, MA: Pearson.
Trochim, W. M. K., & Donnelly, J. P. (2008). The research methods knowledge base. Mason, Ohio, USA: Atomic Dog.