Ipsative modeling
27 views
Skip to first unread message
Keith Lau
Jul 12, 2016, 10:13:34 PM7/12/16
to mirt-package
Hi Phil,
The ipsative modeling known in IRT framework is http://www.ncbi.nlm.nih.gov/pubmed/22733226. I think mirt is capable of this model for limited number of pairwise comparison items because of the demanding multivariate normal integral. It may require user-defined probability function via createItem() to compute the integral . But the user-defined probability function only allows for single item's parameters, my question is how to send (1) the all present estimates of all items and (2) the item indexing of present evaluated to user-defined probability function? Or declaring global variables is also fine. Thanks!
Keith
The ipsative modeling known in IRT framework is http://www.ncbi.nlm.nih.gov/pubmed/22733226. I think mirt is capable of this model for limited number of pairwise comparison items because of the demanding multivariate normal integral. It may require user-defined probability function via createItem() to compute the integral . But the user-defined probability function only allows for single item's parameters, my question is how to send (1) the all present estimates of all items and (2) the item indexing of present evaluated to user-defined probability function? Or declaring global variables is also fine. Thanks!
Keith
Phil Chalmers
Jul 16, 2016, 11:02:23 AM7/16/16
to Keith Lau, mirt-package
Hi Keith,
Seems like an interesting idea, though I'm not overly familiar with ipsative models. I'll have to learn more about them at some time in the future before I can make any real recommendations for tackling them. Thanks for the reference though. 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.
Reply all
Reply to author
Forward
0 new messages