The incremental effects of a cognitive map (CM) and a symbolic code (SC) treatment were evaluated (separately and in combination) relative to a standard modeling treatment (SMT) that generated equivalent demand characteristics. SMT in turn was compared to a no-treatment control condition (NT). In simulated job interviews conducted 3 weeks after treatment," significant gains were attributable to CM on four of six social skills criteria. The SC and combination (CM + SC) treatments produced significant improvement on all dependent variables. Evidence favoring SMT over NT was weak. Cost-benefit implications for the theoretically derived CM and SC treatments are noteworthy. Bandura's (1971) seminal work in observational learning has guided the development of social skills training programs in such diverse areas as assertiveness (e.g.,