Augmenting Assessment with Learning Analytics

Learning analytics as currently deployed has tended to consist of large-scale analyses of available learning process data to provide descriptive or predictive insight into behaviours. What is sometimes missing in this analysis is a connection to human-interpretable, actionable, diagnostic information. To gain traction, learning analytics researchers should work within existing good practice particularly in assessment, where high quality assessments are designed to provide both student and educator with diagnostic or formative feedback. Such a model keeps the human in the analytics design and implementation loop, by supporting student, peer, tutor, and instructor sense-making of assessment data, while adding value from computational analyses.