For my thesis I am conducting my analysis, and am having problems with the second part of my analysis. For my second effect, I am looking at the effect of my IV compliance (yes vs no; categorical) on the DV negative emotions (continuous), moderated by skepticism (continuous).
Can I use a two way anova here? As the construct of skepticism has continuous and therefore not grouping variables.
Thank you so much for your help!
Hello Julia,
Yes. If you're running an anova, enter the continuous IV as a covariate.
The analogous model in regression is likely easier to envision, and incorporation of interaction terms (for you, compliance by skepticism) is straightforward.
Good luck with your work.
Anova with a covariate, ANCOVA, and regression with an interaction term all test the same model, so use whichever one is most familiar to you.
I wonder if the whole approach is meaningful. "Compliance" does not seem to be an independent variable, i.e. subjects are not randomized into these groups. If there is no randomization involved, how should one draw conclusions from the statistical model? It's surely ok to use some kind of linear model (e.g. ANCOVA) to express the mean respone (negative emotions) given complicance and skepticism. But inference from this may be wrong to any extend.
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping
- Maps