EJ
About
- Username
- EJ
- Joined
- Visits
- 2,598
- Last Active
- Roles
- Member, Administrator, Moderator
Comments
-
Hi Thorsten, Thanks for presenting the case so clearly. Yes, I think you are spot on. It is interesting that there is so much evidence against the interaction -- I bet this is because you have 5 levels of disparity. As an aside, this is valuable in…
-
I'll need some time to digest this. Will get back to you later.
-
Hmm OK, that is a rather big effect of leaving out this single participant. Of course, you could also argue "how it is that the p-value only changes from .09 to .14 when I leave out this huge outlier in my relatively small data set?" but t…
-
That's interesting. I'll have to look at this a little later (some deadlines now, remind me if I haven't responded in a week), but some of the discrepancies may be due to violations of assumptions (homogeneity of variances). I'll take a look.
-
OK I confirmed my hunch. If you upgrade to the latest version JASP will automatically recognize the NA and show a "." in its place. Cheers, E.J.
-
Depends on your JASP version. If you use 0.8.4 (the latest one), JASP should automatically recognize the "NA" as missing value (you can set this in the preference menu). If JASP does not recognize the NA as missing value and instead classi…
-
Hi Kasia, That's a remarkable test. It does make sense to me to present the quantile that the patient represents in the control population, and perhaps some uncertainty that comes with that quantile (for instance through bootstrapping or a Bayesian…
-
I think that Richard Morey will have more insightful comments. I'll attend him to your post. E.J.
-
Hi Alon, * Convergence is most likely not a problem for these models. * The Bayesian ANOVA in JASP is simply the Bayesian mixed model. So you should be able to get the same result out of JASP as you get out of BayesFactors. * I find this difference…
-
* yes you should be able to add another column (variable); just double-click the data and use your preferred spreadsheet editor, press save, and the JASP file ought to be updated. * wrt replicating your analyses on a different data set, have you tri…
-
Completely correct! E.J.
-
Hi MSB, The correction for multiplicity is usually always through the prior odds. This is even the case in parameter estimation, where you have to spread out your prior mass across more options (imagine a discrete parameter space with an increasing …
-
Well I am prodding people, as you are prodding me. But it will be more effective if you prod on our GitHub page, because then everybody in the team gets to see it
-
It means that the data have changed the odds in favor of models that include the predictor by a factor of 19. When BFincl = .30, you can interpret this as BFexcl = 1/.30 = 3.33, some change in the odds but nothing to get worked up about. E.J.
-
Hi MSB, * The "U" in BF10,U stands for "uncorrected" (the upcoming version will mention this explicitly in the table footnote) -- so yes, they are the same. * Yes, posterior odds = prior odd * BF * Yes, the correction for multip…
-
Hi Jonas, In principle, the question of the presence of an effect is independent of the strength/size of that effect. I would argue that the prior odds can still be based on outcomes for earlier experiments, but then in terms of Bayes factors that …
-
Hi Jonas, It is indeed a little confusing, because "prior" means different things. On the level of models, the prior p(H1) means "what is the relative plausibility of H1?"; on the level of parameters within a model it means &qu…
-
BF inclusion is similar to a regular BF except that it compares two classes of models, one class with the factor of interest and one without. So you start with prior probabilities on the models; comparing the prior probability with vs without the fa…
-
My advice: 1. Have an online appendix in which you present the annotated JASP output (and/or put the annotated .jasp file on the OSF). 2. In the main text, for ANOVA, use "best model on top" and describe the column of BF01s (this is genera…
-
Hi Jonas, With the BF in hand (say 3 in favor of the null), you need to determine your own prior model odds (say 3 in favor of the alternative). Multiplying these numbers yield the posterior odds (in this case, 3 * 1/3 = 1). Then you can transform …
-
Ah, so with the SD's weighted by N-1 instead of N (https://stats.stackexchange.com/questions/66956/whats-the-difference-between-hedges-g-and-cohens-d)? I think that this kind of bias-correction is a frequentist concern, as least as far as inference…
-
I'll post this on the GutHub page. E.J.
-
Hi Alon, This is more the expertise of Richard Morey, so I'll attend him to this, as well as Quentin Gronau. For what it is worth, I think you are right. The Bayesian ANOVA in JASP is really a Bayesian linear mixed model. Perhaps there is a post on…
-
Hi G, A few remarks: 1. The sequential plot stays at BF=1 until about the 10th participant. I assume the first 10 were all from the same condition? Might be good to state explicitly (JASP assumes the participant came in in the order of the rows). 2…
-
We'll clarify this in the table heading for the next release; I've made it an issue on our GitHub page.
-
U stands for "uncorrected", so it does not include the post-hoc correction term that comes from the prior model probability.
-
Thanks for not letting this go. I'll Email the student to request an update. Cheers, E.J.
-
Dear xxPolyG, Yes, filtering is a common chore and although you can double-click the data and do your filtering in, say, Calc or Excel, this is somewhat tedious. So we are definitely going to add filtering capacity to JASP. When exactly this will h…
-
Hi MMA, Well, the posterior distribution is not exactly a Gaussian (it is close, but for low-N it will have thicker tails). But from the median (=the mean if the distribution is Gaussian) and the 95% CI you can compute the SD. Cheers, E.J.
-
Hi Larry, 1. Yes, the priors are all on delta, the standardized effect size (and sd indicates the uncertainty) 2. We are in the process of eliciting a number of other informed prior distributions. 3. For medium-large and large: interesting question…