Self-Efficacy and Smoking Reexamined. Construct Validity and Clinical Utility

Abstract

Clients in a smoking cessation program rated their confidence (self-efficacy) in their ability to resist smoking at the end of treatment and throughout a 6-month follow-up period. Consistent with previous findings, posttreatment scores significantly predicted subsequent smoking status. When concurrent smoking was partialed out, efficacy remained predictive but to a much lesser degree. The partial correlations also indicated that when pitted against concurrent smoking, confidence scores retain some unique predictive power but do not operate as a powerful mediator of the effects of concurrent behavior. Factor analysis of posttreatment scores revealed that efficacy ratings are primarily unidimensional and not situation specific. Efficacy ratings made 2 months after treatment were quite predictive of future relapse. These analyses suggest that efficacy ratings, although generally not a mediating variable, can be useful predictors of relapse, particularly when assessed during the maintenance phase of treatment. © 1986 American Psychological Association.