Social influence constructs: Inferences from a pain management population

The social influence constructs of perceived practitioner expertness, attractiveness, and trustworthiness were measured in a hospital-based chronic pain management training program. Measurement was taken at the end of each week of the 4 week program, allowing for analysis of trend over time for each of the 3 constructs. Subjects (N = 35) used the Counselor Rating Form-Short version (CRF-S) to rate 3 different practitioners: the Staff Psychologist, Psychology Intern, and Attending Physician. A 3 x 3 x 4 (Practitioner, Characteristic, Week) completely within repeated measures design was used. An increasing linear trend for perceived attractiveness and a decreasing linear trend for perceived expertness were hypothesized. The increasing linear trend for attractiveness was found significant only in the Psychologist's ratings, whereas the Intern's ratings showed an increasing quadratic trend that approached significance and the Physician's showed an essentially flat profile. None of the practitioners' ratings reflected the hypothesized linear decline for expertness. The Psychologist's expertness ratings were significant for a linear increase and approached significance for a quadratic increase, whereas the Intern's and Physician's showed no trend. The significant Practitioner x Characteristic interaction suggested that the levels of Characteristic ratings varied more in relation to the Practitioner being rated than to the Week in the program. Factors associated with the Practitioner, such as title, status, degree, professional discipline (i.e., medicine versus psychology), and work experience seemed to be related to differing levels of expertness, attractiveness, and trustworthiness ratings. Post hoc analyses via stepwise multiple regression showed that weekly CRF-S ratings were useful as predictor variables for three outcome measures. The Psychologist's ratings from Weeks 1 and 2 predicted from 37% to 54% of the variance in two self-reported pain outcome measures. In contrast, the Physician's ratings from Weeks 3 and 4 predicted from 35% to 42% of the variance in two self-reported outcome measures, one for pain intensity and one for satisfaction with services. Theoretical considerations and clinical applications based on the findings are discussed.