Two empirical studies are presented which investigated Sexpert, a rule-based expert system for the treatment of couple sexual dysfunctions. The following hypotheses were evaluated: (1) couples' attitudes towards computerized psychotherapy would improve subsequent to an interaction with Sexpert, (2) couples would find Sexpert more engaging than other sources of information, (3) couples interacting with Sexpert would report learning more about their relationship than those exposed to other sources of information, and (4) Sexpert could influence the sexual behavior of volunteer couples. The research hypotheses were tested in 2 studies (81 & 77 couples respectively) which compared the effects of Sexpert, a sex education book and video, and a therapist, with a no-information control (crossword puzzles). A rigorous statistical analysis of the reliability of couples' self-report of their sexual behavior was made. In study 1, positive attitude shifts towards Sexpert were replicated with semantic differential as well as dissimilarity scales. In study 2, positive attitude shifts were only demonstrated via a ranking methodology. Sexpert was reported to be the most engaging of the experimental manipulations tested in either study. Couples provided with information via Sexpert, video, book, or therapist retrospectively reported significant positive changes in communication, sexuality, feelings, and relationship satisfaction. The highest change ratings were given by participants in the therapist group, followed by the Sexpert, video and book groups. Participants also reported significant learning about many aspects of their relationship and sexuality; however, there were no significant differences between the Sexpert, video, and therapist conditions. The sexual behavior monitoring data did not corroborate retrospectively reported changes in behavior. These inconsistencies in behavioral reporting were attributed to couple reactivity to the monitoring of sexual behavior and to s

No description available.