EJ
About
- Username
- EJ
- Joined
- Visits
- 2,598
- Last Active
- Roles
- Member, Administrator, Moderator
Comments
-
Hi Boo, If the data are exchangeable between pilot, Exp1, and Exp2 (a big if!), then you can just label all of that data as "condition A" and compare it to "condition B" (for an unbalanced test, but that's OK). It is difficult …
-
Hi Jamie This may take a while. The BRMS package may do this, but it is not set up to do Bayes factor tests. I am not aware of any formal developments for Bayes factor MANOVAs, unfortunately. Then again, to paraphrase Lord Rutherford: "if your…
-
It is a matter of taste, imo
-
I'd report both and acknowledge the uncertainty. Generally it is a good idea to include more knowledge. In particular, if the observed interaction is qualitatively consistent with the predicted interaction, this should help the model with the intera…
-
Dear AnnalenaB, Thanks for this interesting and very relevant question. I believe the Bayesian directional test makes complete sense. Suppose we start by testing the point H0 (the skeptics' position: d=0) against the two-sided alternative H1 (the p…
-
Hi Stats, I appreciate the sentiment. To an outsider, this must make a strange impression indeed. Let's just say that, if it had been up to me, this situation would never have arisen, and it took me completely by surprise. For a few years now, brid…
-
Thanks Sebastiaan. Coincidentally, our lead programmer, Bruno Boutin, is French so if there is a desperate need for communication in French we can oblige. Cheers, E.J.
-
Hi Felipe, That suggests to me even more that it is an issue with the way the data are read in. But we'll see I guess. Please keep us posted! E.J.
-
Hi Arran, This is reasonable. Also, this distribution is close to several others that we have elicited from experts. For instance, it is similar to the "Oosterwijk prior" (see the informed t-test paper by Quentin Gronau, Alexander Ly, and…
-
Hi Peter, I'll ask Johnny for more information on these tests. They come from R packages, and I think we've checked them against results from other programs, but Johnny knows more about that. Nice project, by the way! Cheers, E.J.
-
Hi Irepet, What appears to be the case is that every row (subject) shows an effect of about the same size, say like this: 1 200 300 2 700 800 3 400 500 If you drop the repeated measures aspect the results will have sufficient noise/variabil…
-
Hi Stats, Jamovi is a project that was set up by some of the people who were hired to help out with the initial implementation of JASP. The reasons for why Jamovi has started at all are a little mysterious -- I certainly never understood it. You ma…
-
Baby steps! Still very much on our radar though E.J.
-
Hi Boo, I gather that you used the pilot data for the BF t-test for Experiment 1. If you use the updating method, then you ought to use the knowledge after Experiment 1 for the analysis of Experiment 2. This knowledge includes the pilot data. So it…
-
Yes.
-
Hi fcorchs, JASP uses R, so the discrepancy will probably be between R and Python. If you post this issue on our GutHub page then the person responsible can look at the specific code (for details see https://jasp-stats.org/2018/03/29/request-featur…
-
Hi Martin, Ah, this is an analysis I am not expert on. But I do know it is a frequentist analysis. So is the H-measure a Bayesian concept? I will ask Erik-Jan (no, this is not strange). Cheers, E.J.
-
Hi Butler, I would usually report the number (so not BF > x). How you report that large number -- I don't have a preference. Surely the APA offers sage advice on reporting large numbers? I would generally go with what people find easiest to unde…
-
Hi Ayelet, Yes, the subscripts refer to the hypothesis; BF_10 = 3 means the data are three time more likely under H1 than under H0; BF_01 = 2 means the data are twice as likely under H0 than under H1. Adjusting the scaling of the plots: we are wor…
-
I'll pass this on to Johnny. We use a particular R package I think. Cheers, E.J.
-
Hi Siran, Thanks! We don't have improper uniform priors for ANOVA or t-test. You could set the scale of the Cauchy to its maximum (2, I believe); this is so spread out that your results should not differ too much from those of a uniform prior (in …
-
Hi Martin, Can you post a screenshot, so I know exactly what you are referring to? Cheers, E.J.
-
Hi Haver, Sorry for the tardy response. This should not happen, obviously! If you post this issue on our GitHub page (for details see https://jasp-stats.org/2018/03/29/request-feature-report-bug-jasp/) then we can address this issue effectively, C…
-
Hi Mathieu I don't think we have this yet, but it would be an excellent suggestion for our GitHub page! Cheers, E.J.
-
I'll attend Richard Morey to your question. E.J.
-
Hi Dario, "calculate how much the old version of the test is correlated to the new one" Since the old and new test share a lot of items, the correlation has to be high. To compute the correlation, ideally you have the same people answer a…
-
Hi Dario, Well, first off, data from a Likert scale can never be normal, because the scale is discrete. But sum scores across several items (or average scores across participants, or averaged sum scores across participants) can be approximately nor…
-
Hi Shun, Yes you can; it is under the plotting options in the ANOVA menu Cheers, E.J.
-
Hmmm. There are at least two solutions to this problem. The first is simple: just eyeball the posterior distributions of the beta's. Although informative, this is of course not a formal test. The second solution would be to compare the models with t…
-
We've made a lot of progress since then, so I think we are in a good position to address this issue now.