As part of efforts to enhance academic achievement in higher education, incoming first-year students are becoming more and more subjected to surveys and assessments, e.g., regarding motivation and learning strategies. The Learning Strategies and Motivation Questionnaire (LEMO; Donche, Van Petegem, Van de Mosselaer, & Vermunt, 2010) is one of these surveys, applied mostly in professional bachelor programmes. The current study examines the reliability and predictive validity of the LEMO questionnaire in a sample of 416 first-year university students. All 13 scales were included in the study, i.e. Concrete Processing, Analysing, Memorising, Critical Processing, Relating-Structuring, External Regulation, Self-Regulation, Lack of Regulation, Amotivation, Controlled Motivation, Autonomous Motivation, Self-Efficacy, and Learning Together. In line with its reliability in previous studies, Cronbach’s alfa of most LEMO scales was below .70, which is the minimum threshold for scientific research, as was the Composite Reliability of eight of the 13 LEMO-scales. A confirmatory factor analysis showed that several factor loadings were below .70, resulting in an average variance extracted (AVE) below .50 for 11 of the 13 scales. Most scales had no or only a limited correlation to first-year GPA (FYGPA). Only Self-Efficacy and Analysing correlated ≥ .20 with FYGPA. These two scales explained 10.4% of the variance in study success. Hereby, Self-Efficacy is the most important predictor. The other 11 scales had no significant contribution to the prediction of academic performance in addition to Self-Efficacy and Analysing (ΔR2 = 3.4%, n.s.). Additional analyses showed that the correlation between the LEMO scales and FYGPA varied according to Bachelor programme.
Read full abstract