I am planning an experimental research. Normally, in experimental studies, equivalent groups are selected and the effect of the method is investigated by applying different methods. However, my aim is not to reveal the effect of the method applied. My goal is to demonstrate the impact of group diversity. In other words, I will apply the same teaching method to the non-equivalent A and B groups. If there is a difference, I will attribute it to the difference of the group. Do you think my thinking is correct?
Hi Okan, If you find out A and B have different outcomes, it could be because they learn differently from your lesson or it could be because the groups are already different. I feel like your strongest design would be a 2 by 2 with group (A versus B) and condition (experimental versus control with your lesson or filler). Your diversity hypothesis is an interaction between group and condition. You'll also know if your lesson worked (main effect of condition) and if your two groups are different overall (main effect of group). Good luck with your study! ~ Katie
I think there is a misunderstanding here. Diversity is not a dichotomous variable, but a continuous variable. One should therefore use the full information and not artificially reduce it to A and B. Therefore, I do not think an experimental approach is appropriate.
Furthermore, diversity is multidimensional, because one can, in principle, use any variable where interindividual differences may be present as indicators of diversity (IQ, gender, hair color, religion, sexual orientation, body weight, age, and and) .
Let us assume for simplicity that one would consider only one variable, say age. Then one would not take an experimental approach with two groups. One would
1) Apply the treatment to all groups.
2) Assign a value of age diversity to each group (for example, the variance of the group on the age variable).
3) Simply take the age diversity value as a covariate of the treatment effect. Thus, what you do, is a moderator analyis.
It is important to note that the units being statistically tested are not individuals, but groups. They must be available in sufficiently large numbers to provide the necessary test power for the effect of diversity on treatment.
Experimental research can be grouped into two broad categories: true ... which is not really a research design, but an all-inclusive term that includes all types of research ... For example, in order to test the effects of a new drug intended to treat a ... groups are similar (equivalent) to each other and to the control group