Analysing Learner Behaviour through Logfiles for Quality Assurance and Student Evaluation

Learning environments provide extensive logging capabilities that document learner activities. Careful interpretation of these logs can be used for quality assurance and student evaluation. We present a logfile processor that reads Toolbook logfiles and collects data for determining “hot spots” (i.e. particularly good or bad spots) in a course or in learner behaviour. Hot spots show up as extreme values during the analysis; they are candidates to be looked at by instructional design experts. The logfile processor’s methods for data collection can be customised using an event mechanism. Its output — either flat ASCII or structured XML — is further analysed: We present exemplary diagrams based on logfiles from a sample course.