Robustness check Bayesian repeated-measures ANOVA
Hi,
For our study, we did a 2 by 2 Bayesian repeated-measures ANOVA on our data. We wanted to assess the sensitivity to the prior specification (we used the default settings), so we wanted to follow the mixed ANOVA example on https://osf.io/wae57/ by running the same analysis for different options (because the classic robustness plot is not available for Bayesian RM ANOVA). However, since we do not have fixed effects, random effects, nor covariates, I wonder whether this makes any sense?
Thank you!!
Comments
Hi IBR,
I believe you do have fixed effects. If you go to "Additional options" and change the prior width, the BF should change -- right? If fact you can just change each of the three scales by an order of magnitude; if the results remain identical, it is highly likely that this factor is not included in your model (but I'm pretty sure the fixed effects is included)
Cheers,
E.J.