Examining Learning Engagement in Context Using Experience-Sampling Method

• In Event: 61.015 - Embracing and Modeling the Complex Dynamics of Motivation and Engagement: Contextual, Temporal, Dynamic, and Systematic

Abstract

Theoretical Framework and Purpose
This research investigated how contextual factors influence self-efficacy and cognitive engagement from a social cognitive perspective (Bandura, 2001). Cognitive engagement is defined as the learning strategies (deep vs shallow) that learners exhibit (Greene, 2015). The extant research on cognitive engagement is largely based on self-report measures collecting data before or after learning events (Fredricks & McColskey, 2012). These measures ask participants to respond to context-dependent questions when they are not in the context of interest, which can compromise response validity (Sinatra, Heddy, & Lombardi, 2015). Experience-sampling method (ESM) is an alternative data collection technique that enables assessment of learning in context (Csikszentmihalyi & Larson, 1987; Zirkel, Garcia, & Murphy, 2015). ESM with mobile technologies affords a unique opportunity to investigate in-the-moment interactions among personal factors, environment features, and cognitive engagement.
We designed an IOS app that asked participants to set up study events in a calendar, prompted participants to answer ESM-surveys during those study events, and collected participants’ responses and contextual information. We examined how interactions between students’ self-efficacy (both as a course-average perception, and variations in the moment) and contextual features (see figure 1) influence cognitive engagement.
Methods & Data Analysis
Participants included 66 pre-service teachers from a large US university. They completed event-based ESM surveys every day for one week prior to the midterm and final exams.
ESM surveys included measures for deep and shallow learning strategies, and task-specific self-efficacy from Greene and associates (2004). Students also provided contextual information for each study session, including location, study partner, and reason for study (to meet deadlines, for leisure, or as routine). The app collected GPS coordinates that were translated into location names using Google Map services, and used to triangulate with self-reported study locations.
We employed within and between effects decomposition analysis, using hierarchical linear modeling (HLM) to model study sessions within students (Raudenbush & Bryk, 2002). Between effects were modeled using person-level averages of self-efficacy across study occasions. Within effects were modeled using occasion-specific deviations from person-level averages. Thus, between effects represent the association between students’ average self-efficacy for the course and cognitive engagement, whereas within effects represent the association between occasion-specific self-efficacy and engagement.
Results
Results showed that study location significantly moderated the relationship between person-average self-efficacy and shallow engagement (Table 1; p<0.05). Moderation effects of reasons for studying were also found. Overall, results suggested that environmental-driven contextual features, such as location, have an influence on how students’ self-efficacy is associated with shallow engagement. Personal-driven contextual features, such as students studying for leisure or as a routine, have more bearings on how self-efficacy is associated with deep engagement.
Significance
Social cognitive theory suggests that engagement occurs through an interaction among students’ personal characteristics, behaviors, and environmental features. Through ESM supported by mobile technologies, our research is a novel attempt to better understand the dynamic, real-time interactions between these three factors in the learning process. Our findings have important implications for reconceptualizing self-efficacy and cognitive engagement as dependent on contextual variables.

Authors

• Kui Xie, The Ohio State University

• Vanessa Wanchanit Vongkulluksn, The Ohio State University

• Benjamin C. Heddy, University of Oklahoma