[Rasch] One model for three constructs and four groups (was RE: PCA and DIF)

Edward Li edwardfli at hotmail.com
Wed Jun 8 15:16:01 EST 2011


On Tue, 7 Jun 2011 10:05:36 -0500 Stuart Luppescu wrote:
 
"Very often high school teachers score lower than elementary school teachers and show DIF on certain items. If the DIF is really large we sometimes make different versions of some of the items for elementary and HS. But too much of this obscures the differences in the measures. If we make completely different elementary and high school versions of the measures we get much better item fit, but then we can not make comparisons among all schools, since elementary schools will be measured only relative to other elementary schools, and high schools only to other high schools. In this case, the mean item difficulties for elementary schools will be 0, and high schools also 0, so elementary vs. high school differences in teacher and student measures will be obliterated"
 
It might not be the best idea, but I'm wondering if it's possible to split items showing DIF to create two versions and use those items showing good fit and no DIF as link items? In this case, we can link two versions and put elementary and high school measures on the scale before comparing them.
 
Cheers,
 
Edward


 
From: slu at ccsr.uchicago.edu
To: rasch at acer.edu.au
Date: Tue, 7 Jun 2011 10:05:36 -0500
Subject: Re: [Rasch] One model for three constructs and four groups (was RE: PCA and DIF)

On Sun, 2011-06-05 at 19:28 -0700, Juanito Talili wrote:
> My question is: What does it mean when DIF test is significant for
> several items  and nonsignificant for other items?  Can I still use
> the instrument for the four shops in the future ?
 
This is a tricky question. You have to balance the amount of
mis-measurement with your desire to compare measurements across shops.
 
In our work, we survey elementary school and high school teachers and
students. Very often high school teachers score lower than elementary
school teachers and show DIF on certain items. If the DIF is really
large we sometimes make different versions of some of the items for
elementary and HS. But too much of this obscures the differences in the
measures. If we make completely different elementary and high school
versions of the measures we get much better item fit, but then we can
not make comparisons among all schools, since elementary schools will be
measured only relative to other elementary schools, and high schools
only to other high schools. In this case, the mean item difficulties for
elementary schools will be 0, and high schools also 0, so elementary vs.
high school differences in teacher and student measures will be
obliterated. 
 
In the same way, if you let each shop have it's own item difficulties,
you will not be able to measure between-shop differences.
-- 
Stuart Luppescu -=- slu .at. ccsr.uchicago.edu        
University of Chicago -=- CCSR 
才文と智奈美の父 -=-    Kernel 2.6.38-gentoo-r6                
As for elegance, R is refined, tasteful, and
 beautiful. When I grow up, I want to marry R.   
 -- Andy Bunn (in a discussion about whether R is
 simple, powerful and       elegant)       R-help
 (May 2005)  
 
 
 
 

_______________________________________________ Rasch mailing list Rasch at acer.edu.au Unsubscribe: https://mailinglist.acer.edu.au/mailman/options/rasch/edwardfli%40hotmail.com 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20110608/d5bfc0f4/attachment.html 


More information about the Rasch mailing list