Date of Award

8-2026

Degree Name

Doctor of Philosophy

Department

Psychology

First Advisor

Scott Gaynor, Ph.D.

Second Advisor

Amy Naugle, Ph.D.

Third Advisor

Amy Damashek, Ph.D.

Fourth Advisor

Taylor Weststrate, Ph.D.

Keywords

Acceptance and commitment therapy, case formulation, process-based therapy

Abstract

Process-Based Therapy (PBT) is a novel approach to clinical research and practice. PBT is proposed as an alternative model to classic diagnostic frameworks. It emphasizes conceptualizing psychological problems through adaptive and maladaptive processes of change. The aim of current PBT research is to develop reliable methods of organizing change processes into network models. In network models, change processes are presented in nodes connected by lines and arrows to represent the interacting relationships between those processes for a given individual. The present project aims to evaluate the inter-rater reliability and predictive validity of a novel semi-structured interview method to aid in constructing network models.

In the first study, participants are recruited based on endorsed symptoms of depression, anxiety, or attention-deficit/hyperactivity disorder. They take part in the researcher-developed Process-Based Clinical Interview (PBCI). Processes of change from Acceptance and Commitment Therapy (ACT) are used as the pool of targets to assess for each participant in the interview. Individual ratings for the relationships between each pair of ACT processes are generated by the interviewer using participant endorsements regarding those processes. A scoring algorithm for these ratings is used to identify two ACT processes with the strongest hypothesized influence over the participant’s overall network. These processes are identified as ideal treatment targets in therapy. Interviews are audio recorded and independent reviewers use the same PBCI procedures to generate their own ratings based on participant endorsements. The study clinician’s and independent reviewers’ ratings and selected target processes are assessed for inter-rater reliability using intraclass correlation and kappa coefficients. Results show a modest kappa estimate of k = .73 for agreement on the top two selected processes across all reviewers and participants. Agreement indicated by intraclass correlations ranged from good to excellent for the individual PBCI ratings (Range: ICC = .844 - .977). The PBCI procedures overall demonstrate satisfactory inter-rater reliability.

In the second study, ten participants are administered five sessions of ACT. Treatment exclusively targets the processes indicated by each participant’s PBCI results. ACT processes are assessed weekly using validated questionnaires and daily using individual-specific items from those questionnaires. Patterns of change are assessed using gradated reliable change indices to determine how well the individual PBCI ratings predict downstream effects of improving the target processes on any non-target processes. Reliable change criteria on target processes are observed for nine participants. Downstream improvements are observed for eight participants in weekly data and two participants in daily data. Analyses show the PBCI ratings correctly predicted 70.45% of process-specific cases of downstream effects in weekly data and 80.00% in daily data. The results indicate good predictive utility for the PBCI.

The present study builds on prior literature by offering a standardized method of hypothesized network model construction. Limitations include the lack of multiple interviewers conducting the PBCI to control for interview-specific bias and the lack of a control group for the treatment study to control for non-specific therapy effects. Overall, the PBCI shows promise as a process-based assessment tool and warrants future research to refine its methods.

Access Setting

Dissertation-Open Access

Download

Share

COinS