Clamshell item and test information plot

48 views
Skip to first unread message

SKK

unread,
Jul 16, 2017, 5:05:00 PM7/16/17
to mirt-package
Hi everyone, 

I want to create clamshell item information plots however I have no idea about how to do it. If you could share a sample syntax for it, that would be sincerely appreciated. 

Best wishes, 
Seval

Phil Chalmers

unread,
Jul 17, 2017, 2:21:22 PM7/17/17
to SKK, mirt-package
There is a way to do it in the package, but unfortunately it looks to be broken on the current CRAN release (this plot is rarely used). If you install the latest dev version on Github this is patched, and can be run with the following syntax:

mod <- mirt(Science, 2)
plot(mod, type = 'info', degrees = seq(0,90,by=5), npts = 10)

I've attached the output to show how this plot should look.

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

Rplot.png

seval kula

unread,
Jul 17, 2017, 3:04:34 PM7/17/17
to Phil Chalmers, mirt-package
Dear Chalmers, 

Thank you very much for your kind help. I am so sorry but I could not understand what I should install from Github. I am a novice in R. I would appreciate if you could explain in more detail. I installed devtools package, and tried the syntax. It said can't plot high dimensional solutions. 

Sev

Phil Chalmers

unread,
Jul 17, 2017, 3:18:57 PM7/17/17
to seval kula, mirt-package
Does your model have more than 2 dimensions? If so, then you won't be able to create these plots.

Phil

seval kula

unread,
Jul 17, 2017, 3:25:24 PM7/17/17
to Phil Chalmers, mirt-package
I am using bifactor model. Although the scale have seven dimensions, I thought that I could create the plots since bfactor functions estimate only two slope parameters (one for general,one for group factor) for each item. So, items are two dimensiona actually. I tried drop.zeros= TRUE since I thought that it would ignore the zero slopes with that command, but I think it is irrelevant with my situation. 

Reply all
Reply to author
Forward
0 new messages