Self-regulated learning (SRL) is the process by which students set goals, monitor progress, and regulate behavior during learning. SRL is an acquired skill; by distinguishing students who struggle with content knowledge from those who struggle with SRL, suitable support can be provided to each. We present a novel behavioural measure of SRL which quantifies how consistently students perform during adaptive learning technology (ALT)-supported practice. Using parameter estimates of an item response theory model, we simulate students’ expected performance and compare it against their actual performance. Some students perform worse than expected; such inconsistent performance indicates they are struggling to engage in SRL. We apply our analysis to trace data from primary school students (n = 239) practicing math at three adaptive difficulty levels (easy, medium, hard), yielding ability and consistency scores for each student. We find that consistency is moderately correlated with ability (r =.34, p <.001); that consistency and ability significantly predict national standardised math exam scores (R2 =.54); and that students are significantly more consistent in the easy condition. We present consistency as a novel behavioural measure of SRL which complements established metrics, and can help practitioners and researchers to distinguish students who struggle developing content knowledge from those who struggle with SRL.

Contains fulltext : 299078.pdf (Publisher’s version ) (Open Access)