Inconsistent results between classifcal frequency analysis and the Bayelism repeated measured ANOVA
I got very different results from the repeat method ANOVA without and without the Bayesian method. These results are computed with JASP 0.17.1.0, which has taken the random slope mater into account. I wonder how can I understand and report the above results. Thank you so much!
Best Regards
Ruoyi
Comments
Why do you consider these results to be different? Both methods indicate support for the full model over the null model (and any other model, in fact). Note that the principle of marginality dictates that if the interaction is part of the model, so are the constituent main effects (this is an option you can deselect, but it is on by default)
EJ