EJ
About
- Username
- EJ
- Joined
- Visits
- 2,530
- Last Active
- Roles
- Member, Administrator, Moderator
Comments
-
Hi Erin, If you are referring to the standard errors, this is because the ANOVA model uses a pooled error term. It would help to see screenshots indicating the difference, because right now your question is a little abstract :-) Cheers, E.J.
-
I'm with Frantisek here: just be transparent about the number of tests you did. E.J.
-
Thanks -- can you maybe answer your own question, for the benefit of the other users who may hit upon your post? Cheers, E.J.
-
Yes that's correct. Note that the interpretation of the posterior probability depends on the prior probability. So if you want to mention that I would say that the evidence increased the prior probability from X (whatever it was) to almost 100%. Che…
-
Hi Leonardo_C, Yes, you don't want a partial correlation. As of yet, filters are not analysis-specific. I guess you could re-order the data set such that each column is the class of objects you are interested in, but it is suboptimal. Maybe a GitHub…
-
Available here: https://static.jasp-stats.org/Nightlies/
-
Hi Mike, It is a prior, which means that it reflects the relative prior plausibility of the values for effect size. You can use the default settings, but if you know more then you can change the defaults. For instance, you may know the direction of …
-
I don't think this is possible at the moment. Maybe a good feature request for our GitHub page... (https://jasp-stats.org/2018/03/29/request-feature-report-bug-jasp/) Cheers, E.J.
-
Hi Ladislas -- good point, right now the interim information is not directly accessible other than by visually inspecting the plot. If you create a GitHub feature request to present this information in a table I'll take care of things. E.J.
-
Informed priors are more narrow/precise than default priors; this carries over to the posterior. Effect sizes may be smaller or larger with an informed prior depending on the situation, but what happens is that the observed effect size is pulled to …
-
I'll comment on the first analysis: this is fine, except for the formulation "Evidence strongly suggests that the observed data are 15.93 times more likely to occur under a model which contains an interaction between Time and Group." It is…
-
You can compare them using Descriptives (e.g., plotting histograms etc.). For statistical tests: When each row corresponds to one participant/machine/unit, you have a paired-sample t-test (in the case of two columns; otherwise it is a one-way ANOVA)…
-
Hi Chuan-Peng, "Row-column independence" means that there is no association between the rows and the columns; so knowing the value of a column ("women" or "men") does not provide any hint as to the value of the row (&qu…
-
There are multiple complications here. First you have both subjects and clips, so I think this calls for a crossed-random effects analysis. Then you have multiple dependent variables -- we don't have a Bayesian MANOVA yet, but maybe the mixed models…
-
Hi ClintR, What you could do is use any textbook on this topic, and analyse the example data with JASP. Our machine learning module was inspired by James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An introduction to statistical learni…
-
Dear PariShad, I don't think you need to add subjects as a between-subjects factor; the analysis should automatically do this. [As an aside, with only 5 participants, I wonder whether it even makes sense to conduct this analysis at all -- I hope you…
-
Hi Jonathan, This is really frustrating. I do believe we have this fixed, but I am not sure. This is a typical GitHub topic though. I'll forward this one to the team though. Cheers, E.J.
-
Hi Stefan, Good point! We should also present some of that estimation information in a table, btw. I've created a feature request on our GitHub page (in the future you can do this yourself as well; for details see https://jasp-stats.org/2018/03/29/r…
-
I'll ask the network folks
-
Yes they do. Ranks are much more robust in general.
-
Sample size and t-statistic are sufficient, meaning that no additional information is needed and no information is lost through this summary (as far as inference is concerned). With sample size and t-statistic in hand, you can indeed conduct the Bay…
-
The team never sleeps, and informs me the latest development version also should not crash. You can find it here: https://static.jasp-stats.org/Nightlies/
-
Hi Fabrice, Thanks for bringing this to our attention. I will ask the team, but I do believe we have fixed this RM ANOVA bug. We are also fixing the slowness issue as you have seen -- some basic restructuring has taken place. A new version should be…
-
So clicking on the filter icon and then using the litter bin does not work? If that does not work (it does for me) please make an issue on the GitHub page. Cheers, E.J.
-
This happens when the data are more likely under H0 than under any of the competing models. So there is some evidence for the absence of a treatment effect (or whatever effect is under study), although the strength of that evidence need not be stro…
-
That is definitely a bug. Thanks for reporting! Should you find any others, it works best if you create an issue on our GitHub page (for details see https://jasp-stats.org/2018/03/29/request-feature-report-bug-jasp/) Cheers, E.J.
-
Hi Peter, Good point. I noticed that in the Learn Bayes module, this is not present either; I think that would be the place for it. I'll attend Frantisek (who has programmed the relevant part of the Learn Bayes module) to this. Cheers, E.J.
-
Update This is one correct approach I believe: https://research.tilburguniversity.edu/en/publications/bayes-factor-testing-of-equality-and-order-constraints-on-measure We want to implement this in JASP but that project has yet to start. Cheers, E.J.
-
Update This is one correct approach I believe: https://research.tilburguniversity.edu/en/publications/bayes-factor-testing-of-equality-and-order-constraints-on-measure We want to implement this in JASP but that project has yet to start. Cheers, E.J.