EJ
About
- Username
- EJ
- Joined
- Visits
- 2,531
- Last Active
- Roles
- Member, Administrator, Moderator
Comments
-
Dear BrittJane, The model with the highest R2 is *not* the model that predicts best: it is the model with the best fit. This is always the model with all predictors included. So for model selection, R2 is not very informative. Predictive ability is …
-
Dear JenL, Interesting question. You could argue either way I guess. Usually you would consider the two main effects model and judge whether you need the additional interaction term (this is your approach "B"). However, it appears as if th…
-
Hi Mark, When you tick "Posterior Summary", you can select "Best model" -- this should then present the posterior summaries only for the model with 4 predictors. Cheers, E.J.
-
Dear SClausen, Johnny van Doorn has developed several rank-based Bayesian methods, some of which have already been implemented in JASP. We'll implement the others as well, and then take a good look at ANOVA. Cheers, E.J.
-
Hi Jan, Hmm this is a tricky one. I guess it depends on what model you are using for inference. Although I can't imagine that in practice the difference will be very large. Cheers, E.J.
-
I think it's fine -- I am not a Bain expert so it is a little hard for me to grasp that table, but the standard JASP ANOVA seems fine. E.J.
-
Yes, they operate under different assumptions, so will give different answers. In general, Bain is particularly well suited for testing order-restrictions. E.J.
-
Dear rhlamber, Just to get this straight: What exactly do you mean with "BF of comparisons for groups interaction"? You mean "split" by, say, not only gender but also by favorite color, so that you get a matrix of plots? Regardle…
-
I'll pass this comment along
-
Ah, this one then: http://www.ejwagenmakers.com/2017/RouderEtAl2017ANOVAPM.pdf
-
For raw BFs, the scale is asymmetrical: evidence for H0 (say) goes from 0 to 1, whereas evidence for H1 goes from 1 to infinity. So a BF of 1/200 is of the same strength as a BF of 200. If you take the log, evidence for H0 is -x, and the same streng…
-
Dear tracys, We can help you out but our programming team will probably need some more information. In order to help you effectively please post the issue on our GitHub page (for details see https://jasp-stats.org/2018/03/29/request-feature-report-b…
-
Dear Marius, We can help you out but our programming team will probably need some more information. In order to help you effectively please post the issue on our GitHub page (for details see https://jasp-stats.org/2018/03/29/request-feature-report-b…
-
Dear Flaihai, Thanks for bringing this to our attention. We would love to fix this, but out programming team may need some additional information. Also, this is exactly the kind of issue that we like to see on our GitHub page -- that way, the entire…
-
I'll pass this one on to Richard...
-
Hi Major, I've passed this on to our network experts. They are currently on vacation, so the reply might take a little longer than usual. Cheers, E.J.
-
Dear Luke19, Well, that is not supposed to happen :-) We check carefully, but sometimes there's an issue that we learn about only when users such as yourself bring this to our attention. We generally recommend a post to out GitHub page so that the e…
-
Hi Jihene83500, JASP handles this internally and automatically, no need to do any calculation yourself! Cheers, E.J.
-
Hi BrittJane, Yes, it is in the works! We are waiting for some other software developments to pull the trigger on this one. When we do implement mediation, this will likely be based on the blavaan R package, so you could check that out Cheers, E.J.
-
Hi Jan, I don't think this is a difference that affects the conclusion, both BF are astronomical :-) Cheers, E.J.
-
Hi Cesar, Sorry for the tardy response, I was stuck in a French swimming pool for a while. I will pass your question on to our SEM expert. Cheers, E.J.
-
Hi eniseg2, Thanks for the positive comments about the paper! Bayesian assumptions are generally the same as the frequentist ones. Yes, the DV for the stereogram example is log-transformed in order to remove the right-skew. Reviewers asked us to el…
-
Yes, as eniseg2 says! E.J.
-
Also, the %error will give an indication of the variability; if more stable results are desired, you can up the number of samples for the numerical routine under "Advanced options" -- this should bring down the error% (at the cost of takin…
-
Soon...(this year for sure)
-
Could you post this issue on our GitHub page? This would help us a lot! (for details see https://jasp-stats.org/2018/03/29/request-feature-report-bug-jasp/). Cheers, E.J.
-
Dear Bfzldh, If you post this as an issue on our GitHub page we can address it! (for details see https://jasp-stats.org/2018/03/29/request-feature-report-bug-jasp/)). As far as we are concerned, the attractive feature of the GUI is not primarily dat…
-
Hi Jan, The effect is massive, so it is a plausible result; The ANOVA routine in JASP is taken from the BayesFactor package. What this package does is explained in a paper by Rouder et al. 2012 (Journal of Mathematical Psychology) and this one: http…
-
Hi Matue, Yes it is. We are in the process of developing Bayesian procedures to make these assessments (or lessen the importance of violating these assumptions, for instance through analyses based on ranks) E.J.
-
Hi Izymil, Well, there are so many papers that take a t-test to compare the mean of two groups concerning dependent measure X. In the case X is the variance, but that is fine, it could have been anything else. Cheers. E.J.