Identical command mirt(data, 1, "gpcm") leads do different results.

[Gregor Liegl]

Hello everybody,

Does anyone have an idea why it appears that the absolute identical command x<-mirt(data, 1, "gpcm") yields slightly different results regarding model parameter estimations (tries also differ in numer of iterations needed for estimation) every time I run it?

The same data set does not lead to inconsistent results when conducting graded response models.

I already tried it with set.seed(1) but this had no effect.

My data set: 20 variables with different number of response categories (2 to 6), 8000 responders, incomplete block-design: all participants answered at least 3 items, none answered the full item set, more than 3000 persons responded to at least 15  items.


I'm looking forward to your answers!

Thank you in advance for your support
Gregor

[Phil Chalmers]

Hi Gregor, 

That sounds rather strange, and not something I'm able to reproduce on my end (I've attached a test script). If this problem persists, and it wasn't somehow user error in your R script, then let me know and we'll track this issue on Github with something reproducible. Cheers.

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.
For more options, visit https://groups.google.com/d/optout.

[Gregor Liegl]

Hello Phil,

Thank you for the prompt reply. Your data led to the same issue, but this time only with graded response models. But after de- and reinstalling all my R-related stuff, now everything is functioning well again and both GPCM and graded models yield consistent results - fortunalely!

Thanks again
Gregor