RE: [Rasch] Separability of parameters

Snider-Lotz, Tom tsnider-lotz at previsor.com
Thu Sep 28 07:03:36 EST 2006


Mike,

 Hmm . . . as I understand it, most IRT models employ logistic functions
rather than normal-ogive functions because the former are easier to work
with.  Is the use of a normal-ogive model a critical piece of this --
i.e., if all IRFs were parallel, would a 3-parameter IRT model based on
a logistic function exhibit separability?

Tom

-----Original Message-----
From: rasch-bounces at acer.edu.au [mailto:rasch-bounces at acer.edu.au] On
Behalf Of Mike Linacre (RMT)
Sent: Wednesday, September 27, 2006 4:59 PM
To: rasch at acer.edu.au
Subject: Re: [Rasch] Separability of parameters

Tom:

Separability of parameters requires that all the parameters have
sufficient statistics. This is only true of Rasch models. Other models
with parallel IRFs, such as normal-ogive models, do not have
separability.

Mike Linacre


_______________________________________________
Rasch mailing list
Rasch at acer.edu.au
http://mailinglist.acer.edu.au/mailman/listinfo/rasch



More information about the Rasch mailing list