3PL estimation with fixed thetas

[Guy Ashiri-Prossner]

Hello,

we are trying to re-create some functionality that we currently get in IRTPro (which we want to get rid of), but even though trying for some time, we didn't manage to achieve it:

Consider a .csv file (see attached) with the following columns: examinee id, examinee score, then a column for each item with either binary response or NULL to indicate the examinee did not get this item. The score column is some pre-known ability estimate, in theta terms (i.e. on the scale of -3 to 3).

The contents of this file are actually the responses to some experimental items (which still don't have any a,b,c estimates), while the scores are calculated using a different section of the exam (excluded from the file), in which the items already have parameters.

We read this file into a data.frame but we can't figure out how to estimate the 3PL estimates with the thetas fixed to the score column.

In addition (and this might be just us being unfamiliar with the package), we did not find out how to specify prior values.

Any help would be greatly appreciated.

[Phil Chalmers]

On Thu, May 30, 2024 at 3:37 AM Guy Ashiri-Prossner <guy.a...@gmail.com> wrote:

Hello,

we are trying to re-create some functionality that we currently get in IRTPro (which we want to get rid of), but even though trying for some time, we didn't manage to achieve it:

Consider a .csv file (see attached) with the following columns: examinee id, examinee score, then a column for each item with either binary response or NULL to indicate the examinee did not get this item. The score column is some pre-known ability estimate, in theta terms (i.e. on the scale of -3 to 3).

The contents of this file are actually the responses to some experimental items (which still don't have any a,b,c estimates), while the scores are calculated using a different section of the exam (excluded from the file), in which the items already have parameters.

We read this file into a data.frame but we can't figure out how to estimate the 3PL estimates with the thetas fixed to the score column.

This is not something typically supported in the package (how would you get the true thetas? Only simulations can do that....). If you search through the mirt-package thread for joint ML estimation you can find examples of how to do this though using some of the package internals, but in general I'd avise against it and stick with the usual marginal ML estimation as it's more realistic. HTH.

 

In addition (and this might be just us being unfamiliar with the package), we did not find out how to specify prior values.

See help(mirt.model) for the PRIOR keyword. HTH.

Phil

 

Any help would be greatly appreciated.

--
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/e671a3a1-9d81-4221-b197-a644b8365742n%40googlegroups.com.