ISSN 0033-2941 DOI 10.2466/14.09.PR0.112.3.771-787 Psychological Reports: Human Resources & Marketing 2013, 112, 3, 771-787. © Psychological Reports 2013 ACHIEVEMENT MOTIVES, SELF-EFFICACY, ACHIEVEMENT GOALS, AND ACADEMIC ACHIEVEMENT AT MULTIPLE STAGES OF EDUCATION: A LONGITUDINAL ANALYSIS 1 GUNNAR BJØRNEBEKK Norwegian Center for Child Behavioral Development ÅGE DISETH Faculty of Psychology, University of Bergen ROBIN ULRIKSEN Institute for Educational Research, University of Oslo Summary.—The present study investigated the joint effects of achievement motives, self-efficacy, and achievement goals as predictors of subsequent academic achievement among educational science students. A longitudinal research design allowed for measurement of motivational variables at several stages of educa- tion during bachelor courses (subsequent to the introductory courses), firstly by measuring achievement motives, secondly by self-efficacy and achievement goals. Subsequently, students’ academic achievement level was measured at four differ- ent points in time, until they finished the last course for their bachelor degrees. A multivariate path analysis showed consistent relations between the motivational variables. The motive to avoid failure positively predicted the adoption of avoid- ance goals (both mastery and performance) and negatively predicted self-efficacy. Academic achievement was mainly predicted by the motive for success and per- formance-avoidance goals. The path analysis also showed strong relationships between the examination grades at different points in time. Previous research has investigated the hierarchical organization of achievement motives, competence expectancy and achievement goals, and how these motivational variables predict academic achievement (Elliot & Church, 1997). Whereas achievement motives are generally regarded as important for performance, research has shown mixed results regarding the relationship between achievement motives and academic achieve- ment. In fact, academic achievement may be more directly predicted by expectancy variables such as self-efficacy and achievement goals rather than achievement motives (Diseth & Kobbeltvedt, 2010). The present study further investigated these issues by measuring achievement motives prior to the measurement of self-efficacy and achieve- ment goals, and by measuring subsequent academic achievement at several 1 Address correspondence to Gunnar Bjørnebekk, The Norwegian Centre for Child Behavioral Development, Postbox 7053, Majorstuen, 0306 Oslo, Norway or e-mail (gunnar.bjornebekk@ atferdssenteret.no).