Date of Award

8-2026

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School

College of Humanities and Social Sciences

Department/Program

Psychology

Thesis Sponsor/Dissertation Chair/Project Chair

Michael T. Bixter

Committee Member

Cheryl Gray

Committee Member

Manuel Gonzalez

Abstract

This dissertation developed and validated the Multidimensional Role Ambiguity Scale (MRAS), a measure that separates role ambiguity into five subdimensions: goal/expectation/responsibility ambiguity, process ambiguity, priority ambiguity, behavior ambiguity, and performance criterion ambiguity. Although role ambiguity is one of the most studied role stressors in organizational research, it has been measured largely as a single global score, a practice that masks differences among its facets and may explain the inconsistent criterion relationships reported in prior work. Scale development followed the three-step framework of Singh and Rhoads (1991). Thirty items were generated from working definitions grounded in role theory, and fifteen research assistants supported content validation through an item categorization task. After a pilot study (n = 37), the refined scale was administered to a CloudResearch sample of working adults (final N = 291). The study also introduced a parallel open-ended measurement stream: participant narratives were scored against thirty theoretically derived keyword indicators by two large language models, GPT-5.4-nano and Claude Haiku 4.5, and the resulting ratings were submitted to a novel LLM-based confirmatory factor analysis. Across both measurement modalities, a five-factor model fit the data better than a one-factor model, supporting the multidimensional structure. The MRAS showed strong internal consistency, convergent validity with the Rizzo et al. (1970) and Bowling et al. (2017) scales, and discriminant validity with role conflict. Subdimensions carried distinct predictive signatures: performance ambiguity was the dominant predictor of job satisfaction, process ambiguity best predicted self-rated performance, and behavior ambiguity was the only significant predictor of turnover intentions. The MRAS also explained variance in outcomes beyond both established scales. These findings support multidimensional assessment of role ambiguity and demonstrate a practical framework for validating open-ended psychological measures with large language models.

File Format

PDF

Download

Share

COinS