I have a 2X3 design for a DV that's on a ratio scale. One of my IV's is ordinal (the 3). When I use process by Hayes model 1 to analyze it (with one nominal IV), I get a significant interaction, but not when I do an ANOVA. I think maybe the difference is that the ANOVA "sees" the 1-3 variable as nominal. Is there an ANOVA for an ordinal scale in SPSS?
One can run such an anova in SPSS (I do not have SPSS):
https://www.youtube.com/watch?v=G-YexMVQAGc
In attachments a SAS program and the results for such an analysis using SAS.
You can run SAS programs for free on the SAS on demands website. One only needs to register once:
https://www.sas.com/en_us/software/on-demand-for-academics.html
If your DV is on a ratio scale and one of your IV's is ordinal (1-3), why do not you use linear regression since ANOVA ignores order in your IV?
- Badges
- Science topic
- Similar topics
- Cognitive Science
- Thinking