Fusing Multichannel Trace Data to Understand Self-Regulated Learning With Advanced Learning Technologies

• In Event: 72.061 - Measuring Self-Regulated Learning: Issues and Innovations

Abstract

While most theories, models, and frameworks of SRL tend to agree on some common basic assumptions (e.g., students are actively constructing knowledge, contextual factors mediate one’s ability to regulate aspects of learning), they also differ on their view of some fundamental issues regarding the nature of SRL (e.g., aptitude vs event, role of various contextual agents, number and types of processes, specificity of the underlying internal and external mechanisms, explanatory adequacy; see Azevedo et al., 2013; Schunk & Zimmerman, 2011). Futhermore, the widespread use of self-report measures has severly hampered advances in understanding the temporal nature of SRL as a process (Azevedo, in press; Greene & Azevedo, 2010; Winne & Azevedo, in press). Recently methodological advances have been addressed by interdisplinary researchers who use advanced learning technologies (ALTs; multimedia, hypermedia, intelligent tutoring systems, multi-agent systems) as research tools used to collect the temporally unfolding cognitive, affective, metacognitive, and motivational (CAMM) processes during learning and problem solving (see Azevedo & Aleven, 2013a, b for a recent review).

The emerging use of ALTs as research tools allows researchers to collect rich, multi-channel trace data during learning. For example, the temporally unfolding of CAMM processes deployed by students during learning with non-linear, multi-representational, open-ended learning environments, such as planning, deployment of learning strategies, metacognitive monitoring, learning-centered affective states (e.g., confusion), and changes in motivation (e.g., self-efficacy) (Azevedo, Moos, Johnson, & Chauncey, 2010; Greene & Azevedo, 2010; Moos & Marroquin, 2010) can be collected in real-time for either immediate (i.e., to be fed back to the system to provide immediate adaptive instruction) or subsequent (i.e., to fuse data channels to model a complex SRL process) analyses. In addition, ALTs are excellent platforms to study and examine both adaptive and non-adaptive self-regulatory CAMM processes since learners do not always monitor and regulate them effectively (e.g., unable to deploy emotion regulation strategy to deal with confusion induced by a complex diagram).

This presentation will emphasize the methodological and analytical advantages of using multi-channel trace data to examine the complex roles of cognitive, affective, metacognitive, and motivational (CAMM) self-regulatory processes deployed by students during learning with ALTs. I will provide empirical evidence from several multi-channel trace data, including concurrent think-alouds, eye-tracking, note taking and drawing, log-files, facial recognitions, and physiological sensors (e.g., GSR, EEG) to exemplify how these diverse sources of data can be fused to significantly enhance our theoretical models of SRL, methodological approaches, and analytical tools. Lastly, instructional implications for enhancing SRL during learning will be discussed.

Author

• Roger Azevedo, North Carolina State University