Date of Conferral

7-22-2026

Date of Award

July 2026

Degree

Ph.D.

School

Psychology

Advisor

Leslie Barnes-Young

Abstract

The study was conducted to investigate whether statistically significant differences exist in the quality of treatment plans developed by artificial intelligence (AI) platforms versus treatment plans generated by licensed psychologists, with a focus on intervention specificity and clinical appropriateness. Given the rapid proliferation of digital mental health technologies and the persistent shortage of licensed mental health professionals, empirical evaluation of AI-driven clinical tools has become a critical area of inquiry. Grounded in dual process theory (DPT), which conceptualizes decision making as the product of both intuitive and analytical cognitive systems, the study employed a quantitative, between-groups design. Sixty-six licensed psychologists were randomly assigned to develop treatment plans for one of three standardized psychological distress scenarios, representing distinct diagnostic categories. In parallel, ChatGPT was used to generate one treatment plan for each of the same three scenarios. All treatment plans were evaluated using a researcher-developed rubric, constructed in accordance with DSM-5-TR criteria and evidence-based practice guidelines. Inferential statistical analyses were conducted to determine whether the quality of treatment plans differed as a function of their origin, human or AI. Findings from this investigation may inform the ethical integration of AI into clinical workflows and contribute to the development of data-driven mental health policy.

Download

Share

 
COinS