EJ
About
- Username
- EJ
- Joined
- Visits
- 2,531
- Last Active
- Roles
- Member, Administrator, Moderator
Comments
-
Hi Brian As far as projection is concerned, that's the term I use (with Angelika Stefan and Felix Schoenbrodt, who are working on this as well); the method is not often applied (yet). It is basically a Bayesian power analysis "on the fly",…
-
The F and t values depend on sample size, whereas effect size does not. E.J.
-
Dear blin100, The first sequential plot you showed (where the evidence first goes up a lot, and then goes down) is really anomalous: you would get this if the first half of the data showed a massive effect, and the second half showed an equally mass…
-
I'll ask the team! E.J.
-
Yes, in a meta-analysis you've have to use the same effect size measure for studies, or else you're comparing apples and oranges. What is recommended depends somewhat on the designs. For 2x2 contingency tables the log odds ratio is a popular measure…
-
It seems that there is something going on with your data structure that does not allow you to conduct this test. In your interaction, how many participants are there in each cell of the design? E.J.
-
Three days!! That is not normal. Then again, 14 covariates is a lot. I'll pass this on to the team. E.J.
-
Ask away, that's what the forum is for E.J.
-
Ahhh my bad, good to know it's fixed
-
Reporting BF-0 solved the earlier issue. The new release will fix the labeling error, but you now have a way to get to the right number. If I understand correctly, you now report a new issue, which is that there are still "quite different resul…
-
His response: "Sub-analyses are possible with the case filter feature, and running a meta-analysis on subsets of cases. But combining them in forest plots as in the example is not implemented. If I remember correctly I decided to not add that m…
-
I think you might need to email Richard directly. Keep us posted!
-
The next release is a bug-fix release, due out in a few weeks. But if you want to proceed right now you can just take BF-0 for the Wilcoxon: that will give you the right result.
-
I think it is because you do the one-sided test, and the sidedness changes from regular t-test to Wilcoxon. You can see this when you do the two-sided test first; for the Wilcoxon, most mass is on the negative values. So you should compute BF-0 for …
-
I'll ask our expert! E.J.
-
Yes, see also here: https://forum.cogsci.nl/discussion/6582/ancova-in-rct#latest E.J.
-
Hi Tom, I've asked some team members to respond... Cheers, E.J.
-
The complication here is that you have multiple people in each group. I think what you might need is multivariate ordinal logistic regression. https://link.springer.com/article/10.1007/s10260-018-00437-7 https://cran.r-project.org/web/packages/mvord…
-
Hi Whirly123, I'll ask our expert. BF are our goal for the future, but it requires additional work so we wanted to provide the estimation framework now. Yes, the ANOVA is linear, not generalized. It was implemented as a linear mixed model (in the R…
-
Could you send me your jasp file? (EJ.Wagenmakers@gmail.com)
-
Hi Jaiden, Well, the Bayesian one-sided test simply truncates the prior distribution (and therefore the posterior distribution as well) to be on one side of 0. But that still gives a well-defined posterior distribution, and 95% of the central mass o…
-
Lindley (1985) uses the zero, so "0", but a lower-case "o" seems to be more common (https://en.wikipedia.org/wiki/Odds)
-
Well, if you would just report BF_U you would ignore the post-hoc character of the test. So I would report the BF_U but also the posterior probabilities (which are corrected for the post-hoc nature of the test)
-
Hi Nils, It is conceptually cleaner to say that the BF is just the BF -- for an assessment of the evidence does not matter how many hypotheses you are testing. But if you were testing many hypotheses, in a haphazard fashion, this usually indicates t…
-
Not at the moment. We are working on the implementation of a missing data imputation routine, but we won't have that done for a few months. Cheers, E.J.
-
Well there *are* two bounds -- and one of them is at -infinity. I have a footnote on this in some paper, let me check...Yes, see footnote 3 and earlier associated text in https://link.springer.com/article/10.3758/s13423-017-1343-3 A quick Google sea…
-
Good that you got it to work. I'll pass this on to the team.
-
Hi Jaiden, From memory: for the one-sided test, their are two bounds, but one is plus infinity or minus infinity, right? This sounds silly, but then again, if you want inference that makes sense you should become a Bayesian instead ;-) Cheers, E.J.
-
OK, so our mixed modeling expert is not aware of any Bayesian tutorial papers. I guess we'll have to write our own! :-)
-
Well you could do things this way. Beware to average the log BFs, not the raw BFs -- the mean of 3 and 1/3 is not 1. However, when you describe the problem as follows: "A psychology version of the kind of experiments I'm running/analysing would…