mirtCAT custom first item

17 views
Skip to first unread message

Aaron Kaat

unread,
Jul 14, 2017, 2:53:51 PM7/14/17
to mirt-package
So I see on the mirtCAT wiki page a custom next item function that uses person properties to vary the next item selection (http://philchalmers.github.io/mirtCAT/html/sim-unidimensional_custom.html). Is there a way to do that with the first item? For example, starting age 5 and item #1 and age 8 at item 18 --- or using one of the criteria, starting age 5 and the MPWI for theta = -0.5 and starting age 8 at the MPWI for theta = 0.25? Or something like that? It looks like the custom item selection can only affect subsequent items, not the first one, based on person properties. 
Aaron

Phil Chalmers

unread,
Jul 14, 2017, 3:15:05 PM7/14/17
to Aaron Kaat, mirt-package
I treat starting items slightly different for reasons that many of the selection algorithms simply fail without any item response information. That said, the 'start_item' argument sounds like what you are after, so just create a vector of starting items for each individual prior to running mirtCAT() in whatever way you feel is appropriate. HTH.

Phil

On Fri, Jul 14, 2017 at 2:53 PM, Aaron Kaat <aaron...@gmail.com> wrote:
So I see on the mirtCAT wiki page a custom next item function that uses person properties to vary the next item selection (http://philchalmers.github.io/mirtCAT/html/sim-unidimensional_custom.html). Is there a way to do that with the first item? For example, starting age 5 and item #1 and age 8 at item 18 --- or using one of the criteria, starting age 5 and the MPWI for theta = -0.5 and starting age 8 at the MPWI for theta = 0.25? Or something like that? It looks like the custom item selection can only affect subsequent items, not the first one, based on person properties. 
Aaron

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply all
Reply to author
Forward
0 new messages