jamovi

Does anyone have some insight into the kinds of modules that need to be developed? I figured I would ask the community first before trying to make something that might not be useful. I did write a shiny app for IRT and could port it over to jamovi but I'm not sure how useful something like that would be for people. If anyone has ideas or knows of useful R packages let me know.

http://kylehamilton.net/shiny/IRTShiny/

I have had a question on twitter recently. https://twitter.com/CrimHealey/status/9 ... 5338180608
Might be worth asking this question on twitter as well. Good question to ask.

Hi Kyle,
I would very much like an IRT.jmo module, as I appreciated your MAJOR module.
Look after the corner and find a module for Mokken Scale Analysis, Rasch Analysis ... maybe it's too much

Thank a lot.
Cheers, Maurizio.

Hi Kyle,

I'd like to see model selection added to the linear regression module.

Cheers,
Tamera

I think an IRT module would be pretty neat, but I think it's a good idea to also ask this question on twitter/facebook to get a better idea of what people need.

@tminnick Are you talking about the "build in" linear regression analysis or the new GAMLj module? Model selection/comparison is already available in the base linear regression analysis (although we are gonna release an improvement that allows for interactions in the models pretty soon).

I think something like the laavan package what you the shiny for would be good esp the GCA module.

I agree that an IRT module would be great. Also, a fuller SEM module. The CFA tool is great, but limited. Also, a cluster analysis model (K-means, Hierarchical, and Two-stage), would be very cool. Perhaps a multiple imputation module would be good, too, to help handle missing data.

Yes, a fuller SEM would be nice. The ability to examine measurement invariance would be very useful.

One more shout out to add Ordinal alpha and Omega as reliability options I get many requests to calculate these from SPSS users. It is easy enough in R, but the average SPSS user won't use R.

-shawn