Three quantitative studies of gender and identity in psychotherapy consultations

This paper details the application of three distinctive approaches to the analysis of line-by-line scores for themes of gender and identity in recorded psychotherapy consultations conducted by three male analysts with a female patient. The first method involved commonly used statistical comparisons of the frequency with which gender subthemes and allusions to identity appeared in each consultation session. The results of this study indicate three significantly different patterns of gender material in the communications from the patient with each of the analysts who interviewed her--and from each analyst and patient/analyst system as well. Therapist dominance in this area appeared to be quite strong. The second study involved measures of overall informational complexity for various aspects of the gender/identity sequence of communications. Here too, individual differences emerged. They not only add to the evidence for therapist dominance in these protocols, but also provide indications that sessions differ in respect to the extent to which the information they contain is ordered and repetitive or redundant, as compared to disordered, complex, and varied. The third study availed itself of stochastic methods in which the Box-Jenkins models were used to define mathematically and post hoc, the deeper structure of aspects of the vicissitudes of gender/identity expressions in the course of these consultations. The main finding was that change in speaker role in respect to these themes accelerated in response to random interventional shocks to the system that occurred at the time of measurement and inversely with shocks to the system sustained the previous second. This look into the deeper structure of these sequences revealed considerable sensitivity to recent shocks to the patient/therapist system, much underlying instability, and strong tendencies toward establishing stability or equilibrium when the system destabilized. Of note is the finding that all three consultations, by and large, followed the same configuration--itself a rarely seen model. The implications and promise of these methods are discussed.