Electronic Theses and Dissertations

Andrew Snell

Date

2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Psychology

Committee Chair

Jeffrey Berman

Committee Member

Randy G Floyd

Committee Member

Neil E Aronov

Committee Member

Jia Wei Zhang

Abstract

The construct validation process helps clarify the associations among observable indicators and unobservable phenomena—called constructs. Sometimes, factor-analytic data reveal that indicators of supposedly separate constructs are instead manifestations of the same construct. Previous research has shown moderate-to-strong correlations among indicators of several constructs believed to account for differences in psychotherapy outcome (viz., treatment credibility, outcome expectations, the therapeutic alliance, and therapist empathy). One possibility is that research participants who evaluate the psychotherapeutic process do not distinguish these concepts and, instead, provide evaluations based on general sentiments. The present study used a confirmatory-factor-analytic design and data from five psychotherapy process studies to compare various models depicting the latent factors underlying observable indicators of treatment credibility, outcome expectations, the therapeutic alliance, and therapist empathy. The following models were tested and compared: (a) a one-factor model in which credibility, expectations, alliance, and empathy are one construct; (b) a two-factor model in which credibility and expectations are one construct and empathy and alliance are another; (c) a three-factor model in which credibility and expectations are separate constructs and empathy and alliance are one; (d) a three-factor model in which credibility and expectations are one construct and empathy and alliance are separate; and (e) a four-factor model in which credibility, expectations, alliance, and empathy are each separate constructs. Across each dataset, comparison of model-fit indices revealed that the four-factor model had the best fit. Credibility, outcome expectations, the therapeutic alliance, and therapist empathy appear to be separate and distinct constructs, and efforts to combine them seem to not be warranted.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest.

Notes

Open Access

Download

Share

COinS