MSB
About
 Username
 MSB
 Joined
 Visits
 157
 Last Active
 Roles
 Member
 Thanked
 18
Comments

Hi Chris, JASP uses the afex package for (rm)AN(C)OVA. Among other things, afex (1) makes sure factors have effect coding, and (2) used type III errors (whereas summary.aov uses type II), both of which affect estimates and significances in unbalance…

The reason you have and exclusion BF is that you have  under Bayes Factor  marked BF01. If you want the inclusion BF mark BF10. Note that these are reciprocal  BF01 = 1/BF10, and BF_inc = 1/BF_exc.

Generally, yes  but I'm not sure you want to use whichModels = "bottop"  it is advised to stick with the defaults here (whichModels = "withmain") (see here). Also you can't really get a BF for an F test  as BF are always compa…

These look really good! It is okay if I use them for class? Just one note: on the left panel, the gray points aren't exactly the pointnull, they represent the likelihood of the null value on the alternative prior and posterior distributions. You ca…

The posterior distribution(s) does not equal what you wrote, but the posterior odds do. Note sure what you mean by: That's because the approach doesn't permit defining the alternative hypothesis as, specifically, the degree of secondorder uncertain…

But the Bayes factor does not support updating with respect to H and I. Instead, it involves updating with respect to H and J, where J is not the same hypothesis as I The BF is based on model I completely  it simply incorporated information regardi…

Okay, never mind  I got it now  reading the original post more closely ^_^

Looking at the code, I understand that models with interactions are compared only to models with all the main effects. But why is this?

Hi Richard, I think you might have mixed the prior distribution with the prior odds. Say you want to test if there is or isn't a difference between two groups. For this you would construct two models: Model A, in which the difference is nonzero. Mo…

Your conclusions seem correct. One option would be to report these results. Another option might be to fit and test your specified model (one main effect and the interaction) in R: library(BayesFactor) data(puzzles) # example of 2*2 data # This i…

Thanks Richard! This is very helpful!

Can you post a screen grab of the models and effects tables?

As a footnote, the violation of the assumption of Sphericity might also explain why posthoc ttests gave such strong results  as these tests do not have the Sphericity assumption.

Hi Ronen, Looking at the (frequentist) RMAnova table you provided, that little 'a' marking for the "cong" effect (and its interaction with "go_nogo") indicates your data violation of model assumption of Sphericity. This, in comb…

With a large `N`, a small effect (almost 0) might be "nearly significant", but in a Bayesian framework such a small effect with a large N would more indicative of `H0` being true.

Seems like (as an unintended side effect) inferBF also enables ROPE analysis.

If you want some of the actual JASP code: https://github.com/jaspstats/jaspdesktop/tree/development/JASPEngine/JASP/R https://github.com/jaspstats/jaspdesktop/tree/development/JASPEngine/JASPgraphs/R

For many supported models, you can also the tidyverse approach, with the ggeffects package.

Last I checked, all Bayesian analyses (ANOVA and regression) in JASP (which are done with BayesFactor) are actually linermixed models, so unbalanced designs should not be a problem.

You're 100% right  BF10 = 1/BF01 I always prefer to write the actual values so that they are humaninterpolate (which ever is larger than 1). e.g. it's easier to understand that H1 was 4.784 times less likely than H0 (BF01 = 4.784), than that H1 …

I think it is more appropriate to explicitly note that the second "revered" test is posthoc. If I were reviewing the paper, I would ask about this.

This is a question of priors  your priors didn't expect a reverse congruency effect, so you didn't think of testing for one. This is the same as for NHST  if you did a 2tail test, it would be significant, but in the opposite direction of what yo…

Yeah, but is it the logistical next step? (my wife advised against this joke, but I am my own man!)

@EJ is there a timeline for including Bayesian logistic regression to JASP or R BayesFactor?

No particular reason  just found that is makes it easier to do other calculation with it later (which I do sometimes). You can set the print to show the regular BFs: incBF < inclusionBF(BF) print(incBF, logBF = F) Or you can extract t…

M.L., I've previously posted some of my custom code here in the forum (here>>) that includes a function for computing inclusion BFs.

In JASP you can asses BFs from summary stats using the 'Summary Stats' module (in JASP's bar, when 'Common' is, hit the + sign).

Sure, if I my data is dinodistributed that would be a problem. I was thinking more along the lines of skewed data and the like  where deviations from the norm are crucial for errorrate in NHST, but I would suspect less so for BFs.

If you simply calculate the second (replication) BF only on the replication sample, you are using the same priors as used in the original BF, even though these should be updated to the posterior distribution estimated after the first sample. If I u…

But if you implement model restrictions into JASP then I will have learned R for nothing! Thanks E.J.!