This article reports on meta-analyses of the relations of self-efficacy beliefs to academic performance and persistence. Results revealed positive and statistically significant relationships between self-efficacy beliefs and academic performance and persistence outcomes across a wide variety of subjects, experimental designs, and assessment methods. The relationships were found to be heterogeneous across studies, and the variance in reported effect sizes was partially explained by certain study characteristics. Implications for further research and for intervention are discussed. Bandura (1977Bandura ( , 1982Bandura ( , 1986 has developed a theoretical framework that highlights the role of self-referent thought in guiding human action and change. According to this model, behavior changes achieved through such diverse methods as guided exposure, modeling, persuasion, and anxiety reduction, are in part the result of creating or strengthening one's efficacy expectations (i.e., beliefs in one's ability to successfully perform a given behavior). Self-efficacy has been hypothesized to influence choice of behavioral activities, effort expenditure, persistence in the face of obstacles, and task performance.Recently, several theorists have recognized the relevance of self-efficacy theory to the understanding and prediction of career-relevant behaviors, such as vocational choice and academic achievement. In the first theoretical paper on this topic, Hackett and Betz (1981) hypothesized that people's selfefficacy expectations will determine both their range of perceived career and academic options and their persistence and success in chosen options. In subsequent research, undertaken primarily with college students and appearing largely in counseling and vocational psychology journals, researchers have tested these hypotheses regarding choice and performance.

The present study, based on three samples of college students totaling 1,832 participants, resulted in the conclusion that a 5-level response continuum for the short form of the Career Decision Self-Efficacy Scale (CDSE)1 proved at least as reliable and valid as the 10-level continua used in normative studies. Values of coefficient alpha ranged from .78 to .87 for the 5-level continuum, in contrast to .69 to .83 for the 10-level continuum. Criterion-related validity correlations with career indecision and vocational identity were comparable for the two response continua. Validity with respect to the scales of the Career Decision Profile was examined, as was construct validity with respect to measures of hope, goal stability, and positive and negative affect. Overall, the study suggests the psychometric quality of the CDSE when 5-level response continua are used and adds to knowledge of the nomological network of the construct.1

A short-term psychoanalytic counseling model was used to identify process dimensions and client outcomes. Six counselors saw 16 clients over the course of 2 semesters. Clients responded to measures at pre-and posttest and after each session, and counselors filled out measures following each session. Sessions were audio-and videotaped and viewed by trained raters. P-technique factor analysis identified 4 dimensions: Psychoanalytic Technique, Working Alliance, Client Resistance, and Client Transference. Results of hierarchical linear modeling indicated that counselor use of psychoanalytic technique and the working alliance increased steadily across the sessions, whereas client resistance steadily decreased. The alliance dimension also changed in a quadratic fashion. The process dimensions also influenced each other and were related to better client outcomes. Implications are discussed and suggestions for future research are provided.

The authors explored client psychological distress as a variable in career counseling. Perry-two clients in a naturalistic setting were seen for a total of 290 sessions by 21 counselors-in-training. The results indicated that (a) 60% of the clients were psychologically distressed, (b) clients' scores decreased significantly from pretest to posttest on psychological distress variables, (c) significant linear growth occurred in the clients' perception of the working alliance across sessions, and (d) the psychological distress outcome variables were significantly related to both the first session level ofthe working alliance and its linear growth.