Ggum theta estimates

[Kiff Cu]

Hi Dear Chalmers.

I use ideal point model fordata with 15 six point Likert type items (the scale measures cognitive flexibility). 

The attached file contains the results. Everything was ok with the results except the correlation between GGUM model based estimates of ability and alternative dominance models (GRM and GPCM) is almost perfectly negative. I know that it is possible to reaching such kind of result because the probability of endoring an option is defined differently by dominance and ideal point models. But such kind of "perfect" contrast make me to be sceptic on whether or not I made something wrong.  Also, even everything is ok with this result, I couldn't figure out how to interpret this results. In the literature, there are some articles where theta estimates from different models are perfectly (positively) correlated. 

I really appreciate if you (or somebody who have experience on ideal point models) give some suggestions and directions for interpreting the results.

Best regards.

[Phil Chalmers]

I don't have a great answer to this in that I'm less familiar with the application of GGUMs relative to other models, particularly without a reproducible example to play around with. Sorry. Hopefully someone else knows more about this phenomenon than I.

Phil

--
You received this message because you are subscribed to the Google Groups "mirt-package" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mirt-package...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mirt-package/a4e2b712-d0ef-4835-aaa1-138afb13a637%40googlegroups.com.

[Bo]

Hi Kiff, 

The direction of location parameter is GGUM is somewhat arbitary. I think that may be the reason why you see "perfect" negative correlation.