Paper 1 - Summary

Volume 4 Issue 1 November 2006

Applying Multidimensional Item Response Theory Analysis to a Measure of Meta-Perspective Performance
K. Michele Kacmar¹, William L. Farmer², Suzanne Zivnuska³, L.A. Witt⁴
¹ Department of Management and Marketing, University of Alabama, USA
² FedEx Express, Memphis, USA
³ College of Business, California State University, USA
⁴ University of Houston, U.S.A.

The authors introduce a scale to measure meta-perspectives, my view of your view of me, about one's performance in an organizational setting. We elected to focus on employee performance because it is of perennial interest to both academics and practitioners. When applied to the performance appraisal process, meta-perspectives allow us to investigate how employees think their performance is viewed by their supervisors. Since individuals react to others as they think they are seen by them, meta-perspectives enrich our understanding of the relationship effects inherent in the performance appraisal process. To introduce the concept of meta-perspectives we employed multidimensional item response theory (MIRT). We selected this technique due to the desirable properties it possesses such as non-sample specific item parameter estimates. Using MIRT to model our data allowed us to simultaneously estimate dimensionality and item threshold values. MIRT analyses of data collected from 1,255 full-time workers in two

different organizations reveal that the meta-performance items did not lie along a unidimensional continuum as we suspected. Rather, the items fell out on three dimensions: employee perceptions of the supervisor’s view of employee work ethic, work product, and self-regulation. The resulting structure highlights the complexity of the performance appraisal process. The authors offer suggestions for refinement of the scale and future research.

Keywords: Item response theory, scale development

Download FULL PAPER

Back to Contents