Which Student and Instructional Variables Are Most Strongly Related to Academic Motivation in K–12 Education? A Systematic Review of Meta-Analyses

• In Event: What Research Syntheses Tell Us About the Individual, Instructional, and Contextual Correlates to Academic Motivation

Abstract

Background & Aim
Academic motivation relates to important life outcomes, including achievement (e.g., Baranik et al., 2010), educational attainment (Vecchione et al., 2014), and psychological health (Ryan & Deci, 2017). Accordingly, many empirical studies and meta-analyses have examined factors associated with academic motivation constructs, like intrinsic/extrinsic motivation, performance/mastery goals, or values. However, most meta-analyses focus only on a few variables at a time. Some focus on certain student characteristics in relation to their motivation (e.g., Vasconcellos et al., 2020), and others focus on certain instructional variables and how they relate to motivation (e.g., Bardach et al., 2020). As a result, the field does not have a full depiction of the comparative strength of the relations of the relevant variables to students' motivation. Such comparative information is critical to a more in-depth understanding of academic motivation and ways to enhance it, two issues described by Anderman (2020) and Hattie et al. (2020) as fundamental challenges for the field to address. Adressing these challenges, the current article investigates which student and instructional variables are most strongly associated with academic motivation. We also address the methodological strengths and weaknesses of meta-analyses, analyzing the results' quality, heterogeneity, and currency.
Methods
We provide a comprehensive systematic review of meta-analyses on academic motivation following the PRISMA Guidelines (Page et al.,2021). The current review includes 125 meta-analyses with 488 effect sizes investigating variables associated with academic motivation in K-12 education, based on more than 8,839 primary studies and almost 25 million students. We categorized the variables into student variables (Achievement, Socioemotional variables, Background) and instructional variables (Teacher variables, Interventions, and Technology) and compared categories‘ mean effect sizes using three-level meta-analyses.
Findings
Results showed that, overall, student variables (d=0.41) and instructional variables (d=0.40) had medium and similar mean effect sizes. Out of the student variables, socioemotional variables (d=0.57) were more important than achievement (d=0.44) and background (d=0.21). Out of the instructional variables, we found that teacher variables (d=0.59) were more important than interventions (d=0.37) and technology (d =0.36). The data did not allow to examine results for different motivational constructs due to overlapping construct definitions in the meta-analyses. Regarding the methodological quality, we conclude that meta-analyses quality is sufficient to provide estimates of the effect size. Still, the field needs to improve explaining heterogeneity of effect sizes and to update meta-analyses.
Significance
Our work helps researchers find connecting points for their studies and detect potential research gaps that should be addressed, thus laying a foundation for new and exciting research in academic motivation. Overall, our findings can be interpreted as showing the malleability of students' academic motivation, in that it is not determined by their background, and so suggest that interventions to enhance students' motivation and improving teaching quality in school should be the focus of educational practitioners and policymakers. At the same time, this review clearly showed the impact of the different ways in which academic motivation is defined in various meta-analyses on what can be learned from the literature, something the field needs to consider carefully as it moves forward.

Authors

• Thorben Jansen, Leibniz Institute for Science and Mathematics Education

• Jennifer Meyer, Leibniz Institute for Science and Mathematics Education

• Allan L. Wigfield, University of Maryland

• Jens Moeller, University of Kiel