“…Although some empirical support for social influence theory has been generated, findings are inconsistent and there is much confusion about the relation between client-counselor interaction and client change (Corrigan et al, 1980). Problems include (a) validation of the tripartite structure of counselor characteristics; (b) the possibility of a legitimacy effect (Corrigan et al) in which attributions of expertness, attractiveness, and trustworthiness may be based on the socially-sanctioned counselor role, rather than from specific cues displayed by a particular counselor; (c) specification of the power bases (expert, referent, and legitimate) in which counselor characteristics are rooted (Strong & Matross, 1973); (d) failure to distinguish between actual and perceived counselor characteristics (Dell & Schmidt, 1976;Heppner & Heesacker, 1983); (e) focus on client and counselor perceptions of change rather than actual behavioral change (Zamostny, Corrigan, & Eggert, 1981;Heppner & Dixon, 1978), with very few studies (e.g., Heppner & Dixon;Merluzzi, Merluzzi, & Kaul, 1977) measuring client change through observational rather than selfreport measures; and (f) little attention to the influence attempt, although there have been recent studies of the effects of message discrepancy (Claiborn, Crawford, & Hackman, 1983;Claiborn & Dowd, 1985) and level of client disclosure (Corrigan et al, 1980) in client attributions and change.…”