[Rasch] Disordered thresholds and collapsing (on behalf of Anthony James)

Thomas Salzberger thomas.salzberger at gmail.com
Fri Nov 23 00:36:18 EST 2012

---------- Forwarded message ----------
From: Anthony James <luckyantonio2003 at yahoo.com>
Date: 2012/11/22
Subject: Re: Fw: Disordered thresholds and collapsing

 Dear all,
In 3PL-IRT model guessing is an item parameter.
This means that some items are more prone to guessing and some less.
But it seems to be more logical to consider guessing a characteristic
of persons. Some people have more tendency to guess and some less.
How guessing parameter is interpreted in 3PL?
Are person and item parameters adjusted for guessing in 3-PL?
I mean do person, item, and discrimination parameters change if data
are modeled with 3-PL compared with 2PL and 1PL?


More information about the Rasch mailing list