Symposia
Cognitive Science/ Cognitive Processes
David A. Moscovitch, Ph.D. (he/him/his)
Professor of Psychology
University of Waterloo
Waterloo, Ontario, Canada
Terry Cho, B.A. (he/him/his)
Thesis student
University of Waterloo
Waterloo, ON, Canada
Robert Mackey, B.A. (he/him/his)
Thesis student
University of Waterloo
Waterloo, ON, Canada
Sophie M. Kudryk, M.A. (she/her/hers)
Graduate Student
University of Waterloo
Waterloo, ON, Canada
Jolie Ho, M.A., M.S.
Ph.D. Student
University of Waterloo
Waterloo, ON, Canada
According to the Schema-Congruent and Incongruent Learning (SCIL) model (Moscovitch et al., 2023), effective treatment for social anxiety disorder (SAD) should consist of interventions that both weaken negative self-schemas and strengthen positive ones. Prolific participants from North America and UK (N = 300: 35.6% White, 27.1% Asian, 22.4% Black, 14.9% Other; 52.5% male) were randomized into one of three brief online intervention conditions: (a) imagery rescripting (IR) of a negative social autobiographical memory; (b) positive memory self-processing (PMSP) of a positive social autobiographical memory; or (c) a non-memory control condition. Well-validated measures of affect (PANAS) and self-schemas (BCSS) were collected at baseline, after memory retrieval, and at post-intervention. For positive affect (PA), irrespective of SAD symptom levels, PMSP increased PA more than both the control condition, Mdiff = 8.97, p < .001, 95% CI [3.95, 14.00], and the IR condition, Mdiff = 10.45, p < .001, 95% CI [5.25, 15.65], with no difference between IR and control (p = 1.000). For negative affect (NA), PMSP reduced NA compared to IR irrespective of SAD symptoms, Mdiff = -1.63, p = .046, 95% CI [-.02, 3.24], with no differences between IR and control or between PMSP and control (ps > .30). For positive self-schemas (PSS), irrespective of SAD symptom levels, PMSP boosted PSS compared to both control, Mdiff = 2.07, p < .001, 95% CI [.94, 3.19], and IR, Mdiff = 1.95, p < .001, 95% CI [.80, 3.10], with no difference between IR and control (p = 1.000). Finally, for negative self-schemas (NSS), IR reduced NSS compared to both control, Mdiff = -1.37, p = .016, 95% CI [-.19, 2.54], and PMSP, Mdiff = -1.55, p = .005, 95% CI [-.37, 2.74], with no difference between PMSP and control (p = 1.000). The advantage of IR in reducing NSS was especially strong for high SA participants compared to control, Mdiff = -2.03, p = .024, 95% CI [-.20, 3.87], and PMSP, Mdiff = -2.80, p = .001, 95% CI [-.92, 4.69], with no difference between conditions for low SA participants (all ps > .563). Results highlight the specificity and practical utility of brief memory-based interventions for social anxiety.