Student Motivation in Higher Education: Components, Structure, and Relations With Performance and Dropout Intentions

• In Event: Students' Expectancies and Values After the Transition to Higher Education: A Situated Perspective

Abstract

Purpose and Theoretical Framework
Situated expectancy-value theory (Eccles & Wigfield, 2020) proposes that students’ domain-specific expectancies and task values are shaped by developmental processes and context. Educational transitions such as students’ entry into higher education typically result in mean-level motivational changes (Robinson et al., 2019). However, these transitions can also result in structural changes, such as increasing differentiation or a shift in the salience of different expectancy-value constructs (Wigfield, 1994). Thus, we investigated the internal structure of students’ unique motivational components in higher education. Using a bifactor modeling approach, we examined the differentiation of specific value and cost components and their contribution to a general value factor. Further, we investigated the predictive effects of students’ expectancies, general value, as well as specific values and costs—and their interactions (Trautwein et al., 2012)—on performance and dropout intentions.

Methods
We used longitudinal data from German undergraduates (N=1,435) in their first semesters of various majors (e.g., STEM, law). Students reported their motivation at the beginning of the semester and their final examination grades and dropout intentions (Dresel & Grassinger, 2013) at the end. We measured expectancy with existing items (Kosovich et al., 2015). For value, we adapted and extended items from the school-context (Gaspard et al., 2015; Table 1). Analyses of construct validity yielded satisfactory results. For the main analyses, we examined the multidimensional and hierarchical structure of value in a bifactor exploratory structural equation model (B-ESEM; Part et al., 2020). Afterward, we integrated a latent expectancy factor in the model to analyze relations and interactions of all latent factors in the prediction of the academic outcomes.

Results
First, the B-ESEM supported specific value and cost factors as well as a general value factor (Figure 1). Intrinsic value, personal importance, and emotional cost had particularly strong significant factor loadings on the general value factor. Second, the specific value factors (importance of achievement, utility for job, emotional cost, and opportunity cost) significantly predicted students’ end-of-term performance (Table 2). Expectancy, general value, and the specific value factor for emotional cost significantly predicted students’ dropout intentions. Finally, significant latent interactions indicated that cost not only negatively related to indicators of academic success but also decreased the positive predictive effects of expectancy and personal importance on these outcomes (interactions reported in Table 2).

Discussion
Our analyses of student motivation in higher education support the concept of value as a hierarchical and multidimensional construct (Part et al., 2020). The composition of general value revealed salient aspects of student motivation, which are close to students’ identities and emotions (Eccles, 2009; Umarji et al., 2021). We found distinct and incremental predictive effects of general and specific motivational components on students’ end-of-term performance and dropout intentions. Both main effects and interaction effects highlighted the role of cost in hindering academic success (Barron & Hulleman, 2015; Kim et al., 2021). Considering this complexity, a combined approach targeting expectancy, general value, as well as specific values and costs, might be most beneficial in supporting students’ motivation in higher education.

Authors

• Theresa Schnettler, Westfalische Wilhelms-Universitaet Muenster

• Anne Scheunemann, Ruhr-University Bochum

• Lisa Baeulke, University of Tübingen

• Daniel Thies, Ruhr-University Bochum

• Lena Kegel, University of Münster

• Julia Bobe, University of Paderborn

• Markus Dresel, Augsburg University

• Stefan Fries

• Detlev Leutner, University of Duisburg-Essen

• Joachim Wirth, Ruhr University Bochum

• Carola Grunschel, University of Münster