Designs for learning analytics to support information problem solving

Abstract

Learners working on major learning projects, such as an undergraduate thesis, frequently engage in information problem solving (IPS). In round-trip IPS, learners set goals and develop a work plan, search for and filter sources, critically analyze and mine key information, and draft and revise a final product. Information problem solving is a prime site for self-regulated learning (SRL) whereby learners formulate and carry out self-designed experiments to improve IPS skills and expand knowledge about the topic of the learning project. We describe nStudy, a software system developed to gather ambient trace data that operationally define features of IPS and SRL as learners work on learning projects. We illustrate how trace data can be used to promote learners’ (a) understanding of the topic of a learning project and (b) development of IPS by generating learning analytics, guidance in the form of quantitative and qualitative accounts describing information learners work with and operations they apply to information. Three main challenges are addressed: learning how to plan a learning project, expanding knowledge of the topic of a learning project, and benefiting from and productively contributing to peer reviews of draft products. We conjecture about an emerging ecology for IPS in which big data and learning analytics can be major resources for education.