Title:
A Conjoint Unfolding IRT Model Framework for the Analysis of Preference and Response Time Data
A Conjoint Unfolding IRT Model Framework for the Analysis of Preference and Response Time Data
Author(s)
Sparks, Jordan
Advisor(s)
Roberts, James S.
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
The Generalized Graded Unfolding Model (GGUM) is an unfolding item response theory model that produces single-peaked, nonmonotonic item characteristic curves consistent with a proximity-based response process. The model can be applied to binary or graded item responses, or a mixture of the two. This paper proposes a modification to the GGUM estimation procedure, referred to as the Generalized Graded Unfolding Model with Response Times (GGUM-RT), which includes response time as collateral information in estimating model parameters. The inclusion of response time data illustrates how (a proxy for) cognitive processing relates to the latitude of acceptance construct from social judgment theory. It is also demonstrated to improve the precision of model parameter estimates obtained from the standard GGUM.
Sponsor
Date Issued
2023-11-20
Extent
Resource Type
Text
Resource Subtype
Dissertation