Abstract by Christine Partington
Analysis of Therapist Effects
Therapist effects is an attribute of counseling that is difficult to quantify because the direct effect a therapist has on a client is confounded by the complex interactions between therapist and client gender, ethnicity, and age, as well as the therapist’s chosen method and the client’s previous experience. While previous studies have focused on specific aspects of these issues, the overall complexity has been difficult to examine. Recently, data mining has emerged as a new, alternative analysis technique for understanding the complexity of psychotherapy data. Data mining uses various algorithms and statistical techniques to identify patterns and connections in data that would otherwise be difficult to interpret. The current study intends to use data mining and machine learning algorithms, including multilevel modeling and decision trees, through existing packages, such as Stata, as well as custom python scripts, in order to determine true therapist effects on clients as well as the impact of other factors, including type of therapy, number of psychotherapy sessions, and individual therapist methods.