An Empirical Approach to Studying Psychoanalytic Process

This study was a pilot attempt to investigate psychoanalytic process from an empirical perspective. We randomly selected early and late sessions from an audiotaped psychoanalysis (N = 324 sessions) and scored them using a modified form of the Gill-Hoffman (1982) system. We analyzed these scores via Markov models and comparative probability tests. The coding scheme was reliable, although the percentage of interrater agreement for ratings of patient insight was low. Despite this measurement error, the study yielded interesting findings concerning early and late hours. Across early and late sessions, interpretations facilitated patient transference insight, compared to other interventions within the same hours. The patient was somewhat more likely to follow one insight with a second during late hours. There were several other suggestions of longitudinal change toward greater patient mastery. We also explored the interrelationship between type of interpretation and patient productivity of transference insight. The findings indicated that the piloted method represents a useful way to pinpoint theoretically important interactions for empirical investigation.