EJ
About
- Username
- EJ
- Joined
- Visits
- 2,557
- Last Active
- Roles
- Member, Administrator, Moderator
Comments
-
Dear H_K, Right now I think we offer only the generic version. I agree that it would be nice to add other experimental designs. This would be a great feature request on our GitHub page (for details see https://jasp-stats.org/2018/03/29/request-featu…
-
Thanks. We'll have a small release soon (this mainly fixes some installation issues we've been having, and improves things internally) quickly followed by a release with a lot more new functionality. Cheers, E.J.
-
I've just responded (in that right place)
-
Hi Paula, These are two separate issues: "I need help figuring out the new HEAT map function in descriptives (since 0.16). Haven't figured out how to make it work." Can you be more precise? I had trouble understanding it myself when I firs…
-
I'll bring this to Simon's attention again Cheers, E.J.
-
The new signed installer and zip can be found here: https://static.jasp-stats.org/JASP-SKF-0.16.2.1.msi https://static.jasp-stats.org/JASP-SKF-0.16.2.1.zip NB there is basically a small bug in the installer, where the following happens: - install 0.…
-
:-) great
-
We have this in the SKF version, which has not yet been integrated in the public version of JASP. We can send you the latest SKF version when we've done some checks. Please remind me if you don't get it within a week.
-
I'll forward this to our expert. This seems like a bug though, in which case a report on GitHub would be most welcome. But I'll have our expert look at this first. E.J.
-
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.