Can Chi Square be Used for a Mixed Design with Repeated Measures?

Click For Summary
SUMMARY

The discussion centers on the use of Chi Square analysis in a mixed design experiment involving repeated measures, specifically in the context of moral dilemmas. Participants were divided into two mood conditions (positive and neutral) and asked to solve the Footbridge and Trolley Dilemmas. The authors of the referenced study employed Chi Square to analyze responses, focusing on one dilemma at a time, which suggests that Chi Square can be applied to repeated measures when analyzing independent variables separately. The participants' responses were limited to two choices: "appropriate" or "inappropriate".

PREREQUISITES
  • Understanding of mixed design experiments
  • Familiarity with Chi Square analysis
  • Knowledge of repeated measures in statistical testing
  • Basic comprehension of moral dilemma scenarios
NEXT STEPS
  • Research the application of Chi Square in mixed design experiments
  • Learn about repeated measures ANOVA and its differences from Chi Square
  • Study the analysis of moral dilemmas in psychological research
  • Explore statistical software options for conducting Chi Square tests
USEFUL FOR

Researchers, psychology students, and statisticians interested in experimental design and statistical analysis, particularly in the context of moral psychology and mixed methods research.

Math Is Hard
Staff Emeritus
Science Advisor
Gold Member
Messages
4,663
Reaction score
36
My lab partner and I are trying to recreate this experiment:
http://www.blackwell-synergy.com/doi/pdf/10.1111/j.1467-9280.2006.01731.x?
This appears to be a mixed design.
Half the participants were put in a positive mood and half were put into a neutral mood. After the mood induction, each group solved two moral dilemmas (the Footbridge Dilemma and the Trolley Dilemma).
The type of dilemma is one of the independent variables and is a repeated measure because each group has to solve both moral dilemmas.

Now, it seems the authors use a chi square to analyze the results, and my lab partner and I are really struggling because neither of us ever learned to use chi squares (just ANOVA). I called up a friend who is a stats prof and he says that he seems to recall that a chi square can't be used if there are repeated measures in the design.

But somehow the authors are doing it. From the table on the 2nd page of article, it looks like they only analyze one level of the "type of dilemma" IV at a time. (i.e., they just looked only at responses to the Footbridge Dilemma and then looked only at reponses to the Trolley Dilemma).
So, I guess you can use chi square for a design that involves repeasted measures for one of the IVs?
thanks.
 
Last edited by a moderator:
Physics news on Phys.org
oh, and we were planning on only two reponse choices for a dilemma: "appropriate" or "inappropriate", just as the authors did.
 
ok, never mind. we are going to use a different measure for the dependent variable.