Short and long(er)-term returns of self-efficacy’s latent growth: class attendance and longitudinal impact on knowledge, proximal and distal interest

The broad relationship between students' self-efficacy and interest has been highlighted for decades. This, along with the inherently developmental nature of learning, calls for a more thorough examination of the way fluctuations in students' self-efficacy influence their interest in learning within a single course over an academic year and beyond. The present study investigated how the latent growth of students' self-efficacy beliefs over time influenced their interest in a particular subject area, both short-term (3 weeks) and long-term (47 weeks). Data were collected at eight intervals: four times for self-efficacy within one academic year and three times for domain interest during the following academic year, primarily through weekly e-learning activities at a Japanese university. Attendance over three academic semesters, standardised achievement over one academic year (with prior knowledge control), and domain interest were modelled as outcomes in a latent structural equation model that encompassed a latent curve for self-efficacy beliefs throughout the course of study. The distinct contributions of both the baseline and the growth rate of self-efficacy beliefs to the variance in students' domain interest over the short and longer term were confirmed by the study. Over the course of two academic years, it was found that the growth rate of self-efficacy, relative to its baseline, became increasingly more important for key learning outcomes such as students' domain interest.