Poster #74 - Is School Engagement Distinct from Motivation and Self-Regulated Learning? A Psychometric Evaluation

• In Event: 1-169 - Poster Session 04
  In Poster Session: PS 04 Section - Education, Schooling

Integrative Statement

School engagement has been one of the hottest topics in educational psychology and learning science. The measurement of engagement in extant research was heavily based on Fredricks, Blumenfeld, and Paris (2004), which conceptualized engagement as a construct with three dimensions: (1) Behavioral engagement refers to positive behaviors such as paying attention, and completing assignments; (2) cognitive engagement refers to willingness to exert effort to understand content and work through difficult problems; (3) emotional engagement refers to feelings of connection to content, interest in material, and enjoyment of learning. Two additional components were social engagement, defined as students’ day-to-day school-related social exchanges with peers (Rimm-Kaufman et al., 2014), and agentic engagement, defined as students’ proactive contribution to instruction (Reeve & Tseng, 2011). Despite the large amount of empirical research on engagement as a latent construct, there is no theoretical foundation to support the frequently-used three-dimensional model by Fredricks and colleagues (Boekaerts, 2016). Moreover, the conceptual similarities between engagement and other established constructs such as motivation and self-regulated learning (Betts, 2012; Betts et al., 2010; Boekaerts, 2016; Eccles, 2016; Fredricks et al., 2004) were not adequately tested. Therefore, the purpose of the study is to examine the dimensionality of engagement and whether engagement can be differentiated from motivation and self-regulated learning.

Participants were 839 college freshmen (60% minority; 52% women) recruited from mathematics and science courses in a medium-sized state university in the U.S. South Atlantic region. They completed an online survey that included the 33-item Math and Science Engagement Scales (Wang et al., 2016), which measure behavioral, cognitive, emotional, and social engagement, the 5-item Agentic Engagement Scale (Reeve & Tseng, 2011), and five subscales of the Motivated Strategies for Learning Questionnaire (Pintrich et al., 1993) that conceptually overlap with engagement dimensions. To examine the dimensionality of engagement, the full sample was randomly split into two subsamples. Exploratory factor analysis of the first subsample identified three distinct factors: behavioral-cognitive engagement, social engagement, and emotional engagement. This three-factor model achieved good model fit based on confirmatory factor analysis (CFA) of the second subsample. To determine whether engagement can be differentiated from motivation and self-regulated learning, CFA-based multi-trait multi-method analyses (Eid et al., 2008) were conducted on the full sample. Results indicated that the three engagement factors were moderately correlated with motivation and self-regulated subscales, supporting the distinctness of engagement as a latent construct.

Using a large and diverse sample of college students in mathematics and science courses, this study found a three-factor model of engagement. Although behavioral and cognitive engagement cannot be differentiated from each other, social and emotional engagement factors were found to be distinct. This three-factor model should be further tested and its predictive validity should be examined in future research. This study also provided empirical evidence that engagement is a meaningful construct that is distinct from motivation and self-regulated learning. Such finding may facilitate the theorization of engagement, development and refinement of engagement measures, and further empirical research on the role of engagement in learning.

Author

• Shuyan Sun, University of Maryland Baltimore County
  Presenting Author