As reported here (39 page PDF), "since the introduction of self-report surveys in Self-Regulated Learning (SRL), researchers have shown that self-reports are not a suitable approach to reveal actual SRL processes." So what to do instead? This article looks at "trace-based methods that rely upon learner trace data captured in technology-enhanced learning environments" but warns "it is critical to ensure that the researchers' interpretations grounded in trace data are valid." What does that mean? "There are three commonly considered types of validity in the literature: content validity, criterion-related validity, and construct validity." This report opts for the third form, offering a combination of a data-driven and theory-driven approach; "researchers may begin their analytical procedures by looking at SRL theoretical assumptions, and then use these assumptions to guide systematic identification of SRL processes from trace data." First-rate paper.
Today: 4 Total: 16 [Direct link] [Share]
