Hello Taha,

It sounds as if you intend to use factorial anova as the primary framework for your analysis. Do note that either MLR or a path model approach (simplest case of structural equation models) could be used as well. All this presumes that your data conform reasonably well to the associated assumptions required for either of the three methods.

Your question appears to hinge on how to treat the "two years" aspect of the study. If the very same cases were used each time, and the very same conditions were applied, and the very same outcome measure was used to collect scores, then you could treat the two-years dimension as a repeated measures factor, which would yield a variant of the split-plot design (which appears to be a three between subjects factor and one within subjects factor design).

The SAS manual addresses the split plot set up pretty clearly: https://support.sas.com/documentation/cdl/en/statug/63347/HTML/default/viewer.htm#statug_mixed_sect033.htm

Goo luck with your work.

Hello, Dr. Morse

Your opinion is exactly right, but my doubt is that my research was done in the form of a factorial design in a completely random format in a vase with completely equal conditions in two years.

Now I did the analysis as a split plot and considered the year as a split factor.

But for this reason, I doubted whether my method of analysis was correct or not.