Developing ideal intermediate items for the ideal point model

The importance of intermediate items has been overlooked for years ever since the prevalence of using dominance-based Likert scales in measuring personality and vocational interests. Rather, intermediate items were usually discarded in dominance scale constructions as they were believed to perform poorly in item discrimination (Chernyshenko, Stark, Drasgow, & Roberts, 2007). The current study aims to recognize the importance of intermediate items by showing that they can be perfectly calibrated by the ideal point model. 355 college students were selected to answer a series of personality and vocational interest measures including self-developed intermediate items. Results showed that personality and vocational interest scales demonstrated satisfactory model fits to the ideal point model, but not to the dominance model. Intermediate items also provided more information than extreme items did for respondents with extreme latent traits. Among the four domains (Frequency, Average, Condition, and Transition, “FACT”) of constructing intermediate items, the Average domain was found to exhibit best performance in showing unfolding curves. The possibility of utilizing the results of this study to develop a guideline of writing intermediate items, as well as to construct computerized adaptive tests based on the ideal point model, was also discussed in the paper.

Name thesis:degree_name

M.A.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Psychology

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2013

Author dc:creator

• Cao, Mengyang

Contributors dc:contributor

• Drasgow, Fritz

• ideal point model
• intermediate items
• item response theory
• personality measures
• vocational interest measures

Statement dc:rights

• Copyright 2013 Mengyang Cao

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/44407

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/44407