Sequential analysis and robustness check graph for ANOVA.

kev_bague

Hello,

I'm conducting a Bayesian repeated measures ANOVA with JASP. My design includes 2 predictors (one between subjects).

Is it possible to compute sequential analysis, as it's possible for bayesian t-test. In other words, how compute graph illustrating sequential analysis and bayes robustness check for ANOVA with JASP ?

If it's not possible, do you have tuto to compute it with R?

Thanks for your help

EJ

Hi kev_bague,

JASP does not (yet) offer sequential analyses for ANOVA; the reason is that there are several BFs in play, and only one may be of interest. I guess we could show them all or ask the user to define the one that's of interest. Also, the repeated measures ANOVA can be slow, but that's not a compelling reason.

However, you can easily do a sequential analysis yourself: just execute a new analysis (and store the interim result) after every new participant. At the moment, this is tedious in JASP, but easy in R (I'd use a for loop, but then again, I am getting old)

Cheers,

E.J.

kev_bague

Hi EJ,

Thanks for your answer and for your articles which help me to conduct analysis with Bayesian statistics.

I have another questions about bayesian ANOVA. How can I analyse simple effect ?

Cheers,

Kevin Bague

EJ

Hi Kevin,

I have not looked into this -- in general, it would be worthwhile to develop Bayesian equivalents for all those variants of frequentist ANOVA...but we are not quite there yet.

Cheers,

E.J.