Abstract: There is no simple and straightforward definition of the construct of self-regulated learning. Theorists in educational psychology have narrowed the scope of students' capability to self-regulate through a focus on the academic side of education, namely on learning and achievement goals. However, the messy world of classroom learning creates a situation in which different goals compete for students' attention. Boekaerts' dual processing self-regulation model describes how learning goals interact with well-being goals. We propose that when students have access to well-refined volitional strategies manifested as good work habits, they are more likely to invest effort in learning and get off the well-being track when a stressor blocks learning. Shifting definitions of SRL have led to changing measurement procedures; researchers moved away from decontextualised measures of SRL to domain-specific measures and then on to context-sensitive measures. The validity and reliability of the first generation of SR assessment has been limited and several issues remain. Recently, researchers have designed assessment packages including new instruments that better capture self-regulation as a process (including for example traces of mental events, situational manipulations, and records of student work strategies). A combination of instruments is preferable over a single instrument for assessing self-regulation as a process and the effects of interventions to improve students' self-regulatory capacity. At present, many sound SRL interventions exist and some general lessons can be learned about classroom intervention research.
