The nomological network of self-efficacy and psychometric properties of its general and specific measures

Abstract

Since its proposal in 1977, self-efficacy (SE) has been applied to almost every behavioral undertaking imaginable. Over 30,000 studies have been conducted on SE since its introduction in 1977, and even meta-analyses exist in abundance. Unfortunately, the self-efficacy literature tends to suffer from several common oversights: 1) neglecting measurement properties of self-efficacy scales; 2) inappropriate compartmentalization of self-efficacy by domain; and 3) inappropriate categorization of criteria/outcomes of interest. Accordingly, the goal of the present research was to address the criticisms raised above through meta-analyses of five distinct areas: 1) the reliability of scores from SE scales; 2) the convergence of SE scales within and across behavioral domains; 3) the potentially differential relationships between SE scales and personality traits; 4) the potentially differential correlations between SE scales and cognitive ability; and 5) the potentially differential correlations between SE scales and outcomes. General and specific SE scales were examined for potentially differing relationships with variables of interest. Scales of self-efficacy exceeded basic standards of internal consistency reliability (though these scales were most consistent when at least 5 - 8 items in length) and displayed strong relationships with one another, even at differing levels of specificity and across behavioral domains. Additionally, self-efficacy scales demonstrated similar patterns of relationships with personality across domains. While measures of self-efficacy displayed more variable patterns of relationships with specific criteria, most scales - even those not tailored for the specific criterion - still functioned as acceptable predictors of academic and organizational performance.