[Rasch] Use of misfit as information and logits in MRC

Fernando Soares fsoares at fit.edu
Thu Feb 26 05:49:10 EST 2009


 

After examining a few articles and scanning over books about Rasch and DIF
analysis I realized I might need personal assistance to address my
questions. Is there a professor based in Florida who could possibly advise
me in regards to my analysis strategy, and perhaps who could be interested
in serve as the outside member in my dissertation committee? For your
information, I am located in Melbourne, FL at the Florida Institute of
Technology.

 

Respectfully,

Fernando Soares

 

From: Mike Linacre (RMT) [mailto:rmt at rasch.org] 
Sent: Thursday, January 15, 2009 8:29 PM
To: rasch at acer.edu.au
Subject: Re: [Rasch] Use of misfit as information and logits in MRC

 

Thank you for your question, Fernando.

Kaiser & keller use both demographic groupings and latent classes to
investigate item difficulty (called "behavior difficulty" in  the paper
because the items relate to behaviors).

Your interest appears to be in demographic groups. If so, then a standard
Rasch analysis combined with a DIF analysis (which produces DIF size in
logits) will produce the findings that you are looking for. A Mixed Rasch
model is not needed. You can then use conventional statistical techniques
(such as regression) to investigate the DIF logits.

Does this help?

Cordially,
Mike L.


At 1/15/2009, you wrote:



I'm interested in an application of Rasch analysis similar to:
Kaiser, F. G., & Keller, C. (2001). Disclosing Situational Constraints to
Ecological Behavior: A Confirmatory Application of the Mixed Rasch Model.
European Journal of Psychological Assessment, 17(3), 212-221.

Mike Linacre
Editor, Rasch Measurement Transactions
rmt at rasch.org www.rasch.org/rmt/ Latest RMT:  22:2 Autumn 2008

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20090225/ede634db/attachment.html 


More information about the Rasch mailing list