Export/Save function for factor scores in EFA and PCA

[annamwolf]

Having done an exploratory factor analysis, many reseachers want to continue their analyses (t-test, ANOVA etc.) with the factor scores that can be calculated from the EFA analysis. Afaik, most people use the Thurstone/regression approach, but it would be optimal if you could implement a choice (https://www.personality-project.org/r/html/factor.scores.html).

• Enhancement: Adds the missing feature to purposefully use the results from an EFA and not only see which items create a factor.
• Purpose: Enables people to continue a statistical analysis with the factor scores as, for example, a dependent variable.
  Thanks a lot!

[boutinb]

This kind of feature will be definitely implemented. Hopefully in few months.

[tkasci]

I would also be happy to see this feature in future releases.

[EJWagenmakers]

This should be easier to implement now that we have already included similar functionality for, say, the Machine Learning module.

[patc3]

FWIW I too would like it.. I resorted to reproducing the EFA in R (using psych::fa) and requesting factor scores (and comparing the outputs of JASP and R to make sure I was reproducing it correctly)

If I can help to add this feature please let me know how

[TZaks]

Is there an update on this? I am also in the position of having used JASP to estimate the factors, but need the scores for further analyses.

[EJWagenmakers]

@vankesteren informed me the code does exist:
EFA: https://github.com/jasp-stats/jaspFactor/blob/389a04f7694b535774fd36b6eacad765d62e4a00/R/exploratoryfactoranalysis.R#L578
PCA: https://github.com/jasp-stats/jaspFactor/blob/389a04f7694b535774fd36b6eacad765d62e4a00/R/principalcomponentanalysis.R#L469

Will it be available in the next version, @JorisGoosen @vandenman

[vandenman]

The code exists but it's currently not possible to export more than one column to the data set. As a temporary workaround, we could add a button that exports the factor scores as a csv or so.

[TZaks]

That would be very helpful, if possible. Do we know the time frame? Best, Tara

…

On Mon, 19 Apr 2021 at 11:36, Don van den Bergh ***@***.***> wrote: The code exists but it's currently not possible to export more than one column to the data set. As a temporary workaround, we could add a button that exports the factor scores as a csv or so. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#101 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AQCZVEE3YYQMHZ2ENAVPS5LTJQBUHANCNFSM4GDQKTZA> .

[MCube78]

Hi Everyone, it's been over 5 years since the original request. I hate asking, but I still don't see it in the GUI (am I missing it?). Any workaround yet (like exporting to csv or txt)?

[juliuspfadt]

I am sorry for the delay, this is still on our radar, but not the highest on the priority list. This is internally related to adding computed columns from R (https://github.com/jasp-stats/INTERNAL-jasp/issues/694). Maybe @JorisGoosen can say more about this

[JorisGoosen]

Currently I am rewriting the data-backend, which will allow analyses to easily add columns to the dataset.
Seeing as how the code is already there it should be easy to take those changes by Erik-Jan and make them functional.

[micahgallen]

Since we got the new data editing, I guess that is the first push of the new back end? Sounds like this issue may finally be resolved in the near future - that would be great!!

[JorisGoosen]

Hehe, im working on it right now ;)

[micahgallen]

You guys rule. I've converted many people to JASP and I won't stop any time soon! Love the work you all do. Thanks!!!

[patc3]

You guys rule. I've converted many people to JASP and I won't stop any time soon! Love the work you all do. Thanks!!!

hahaha me too I've converted a ton of people (profs and students alike)

[tomtomme]

related: #2265 - wants this for CFA and SEM

[tomtomme]

@annamwolf
This is now actively being worked on. However it will likely not make it into 0.19 which will be released in some weeks.
But 0.20 might ship this important feature.

[tomtomme]

@juliuspfadt
Nice!
Will this be in 0.19 or 0.20 now?

[koenderks]

It's in the codebase so it will probably be shipped in 0.19.