Hi all,
What is an appropriate analysis technique when you have a 2 (treatment, control) x 3 (neutral, negative, baseline) design and are also interested in the possible effect of a covariate? Originally, I was using ANCOVA but I want to test whether there are interactions between the covariate (continuous variable) and the group.
Hello Kenneth,
1) Use a regression model and include the desired interaction term.
(e.g., Estimated_Y = b0 + b1*C + b2*Grp + b3*C*Grp + ...)
2) Use a path model and build in the desired interaction term as a predictor of Y.
3) It's possible that Andrew Hayes' PROCESS macro (freely available for spss, sas, and I believe R) would handle this design. Essentially, you're asking if C moderates the relationship of IV with Y.
4) Use one-between, one-within anova, and modify the default model ("full factorial") to include a group X covariate interaction. You can see illustrations of this method when people explain how to test the equal slopes assumption of ancova.
Good luck with your work.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress