Self-reports and learning analytics inform study of emotions and relation to behaviors during learning

• In Event: A Comprehensive Look at Emotion Formation and Regulation Across Different Contexts With Various Methodologies

Abstract

Theoretical Framework
Historically, there are two main approaches to understanding self-regulated learning (SRL: Greene & Azevedo, 2007; Winne & Perry, 2000). The first approach relies on self-report measures emphasizing the learner's subjective experience, including internal states, emotions, motivation, and self-beliefs (e.g., Authors, Year; Pintrich, 2000). The second approach relies on observations or performance data extracted from external actions, collecting learners’ interactions across digital platforms and behaviors that learners engage in as they regulate their learning (Authors, Year; Greene, 2018; Winne & Hadwin, 1998; Winne & Perry, 2000). In the current study, we triangulate across the measurement approaches to observe core SRL processes involving knowledge, awareness, and regulation of cognition, behavior, motivation and emotion (e.g., Pintrich, 2004). We do so to develop a nuanced understanding of SRL phenomena related to learners’ emotions and self-perceptions of knowledge about the learning process (i.e., their metacognition and metaemotions; Authors, Year).
Methods and Data
First-year undergraduate students (N = 191; Mean age 24.20, SD = 4.94; 82.2% female) consented to observation of behaviors on a course Moodle site and self-reports.
Different SRL processes are aptly captured through self-report measures, emphasizing the learner's subjective experience, which also include reference to emotional aspects such as emotion SRL (suppression and reappraisal), metacognition, metaemotion, engagement (cognitive, emotional, behavioral), and expectancy-value motivation (Authors, Year; Pintrich, 1991). Cronbach’s alphas showed good reliability (.90>α>.74). Liking, burnout, stress and lack of desire were measured using single item indicators. We measured expectancy-value and engagement, but for this symposium focusing on the emotion components.
Data traces required data collection from performance data that assessed external actions and interactions of students using the digital platform system (Authors, Year; Greene, 2018; Greene & Azevedo, 2010; Winne & Perry, 2000). Behavioral data was logged and labeled with a shared event-naming schema and the related lesson number when applicable and were categorized into seven types: content knowledge acquisition (CKA); course information gathering (CIG); diligence in keeping up with course information (DKC); metacognitive monitoring (MM); self-testing (ST); submitting assigned work (SAW); task knowledge acquisition (TKA).
Results
Self-report and behavioral trace data correlations showed that liking, stress, lacking desire, suppression SRL and emotional engagement were associated with behavioral traces; metaemotion and reappraisal were not. Reports of additional variable- and person-centered analyses of self-reported profiles and behavior patterns and their association will augment these initial findings.
Scholarly Significance
By incorporating self-report and behavioral measurement, we assessed subjective experiences involving academic emotions and found that emotional experiences, beliefs and knowledge, associated with data traces reflecting cognitive-behavioral processes. Findings support theorized relations between specific emotions and forms of cognitive engagement; however, hypothesized relations involving metaemotion and reappraisal were not observed. Association of suppression with task and content knowledge acquisition behaviors emerged as an important, novel suppression-emotion SRL process warranting closer exploration that may contribute to learning behaviors, perhaps by mitigating intrusive emotional responses during challenging tasks.

Authors

• Shani Rosengarten, University of Haifa

• Robert D Plumley, University of North Carolina - Chapel Hill

• Christina Hollander-Blackmon, University of North Carolina - Chapel Hill

• Leiming Ding, University of North Carolina - Chapel Hill

• Alla Hemi, University of Haifa

• Hagit Frenkel Mullerad, University of Haifa

• Matthew L. Bernacki, University of North Carolina - Chapel Hill

• Adar Ben-Eliyahu, University of Haifa