Date of Degree

9-2019

Document Type

Dissertation

Degree Name

Ph.D.

Program

Psychology

Advisor

Loren Naidoo

Advisor

Charles Scherbaum

Committee Members

Harold Goldstein

Yochi Cohen-Charash

Anthony Boyce

Subject Categories

Industrial and Organizational Psychology | Quantitative Psychology

Keywords

work engagement, measurement, ideal point process, item response process, validation

Abstract

Work engagement has been an extremely popular area of research and practice over the past two decades. However, organizational scholars have yet to thoughtfully consider alternative and potentially more appropriate ways of modeling how individuals report their work engagement and, relatedly, measuring the construct. This dissertation seeks to establish and support the position that (1) individuals use an ideal point (vs. dominance) process to identify how engaged they are and respond to work engagement items, and (2) an ideal point framework can be used to develop a construct valid work engagement scale with good psychometric properties. Since no such scale exists, Study 1 details the construction of a new ideal point work engagement scale. That study also documents how this new scale performs against a new dominance scale constructed alongside it in terms of model fit, test information, and score differences. It was subsequently found that a work engagement scale could be successfully constructed using an ideal point conceptualization and methodological approach, and that this resulted in a better fitting scale than the use of a more traditional dominance approach. Study 2 took this a step back to compare the performance of ideal point and dominance models to response data from several extant dominance-style work engagement scales used in the academic literature to see if my arguments hold for measures previously constructed using traditional methods. Comparisons were also made between the rank-order of scores from these scales and the new ideal point scale. Model fit results from this study were mixed where the ideal point process was supported for some scales but not others, and that these differences may be due to the response scale used. Scoring comparisons demonstrated large differences in how the dominance-style scales rank-ordered participants relative to the ideal point scale. Where there were rank-order differences, participants scoring more extremely on the dominance-style scales tended be scored as more moderate on the ideal point scale. Finally, Study 3 was designed to comprehensively investigate the construct validity of the new ideal point scale constructed in Study 1 including qualitatively compare these validation results with those demonstrated by extant work engagement scales. This was done to support the use of the new scale in future research as well as demonstrate that the use of ideal point methodology does not adversely the validity of work engagement scores. Theoretical and practical implications as well as future avenues of research to expand this area of inquiry are discussed.

Comments

A revised version was uploaded on May 5, 2020 with the approval of the dissertation committee chair and the Graduate Center.

Share

COinS