SEED projects 2019 – A new Canvas for teaching and learning

Auto-text analysis for just-in-time teaching

Dr Kaitlin Beare (Faculty of Science)

Automated text analysis is increasingly being used to help bridge the gap between an ideal situation where students write often and the reality of teaching in large classes. In this project we explored automatic text analysis in two interrelated contexts, firstly to give feedback to instructors to facilitate Just-in-time teaching approaches and secondly, to give feedback to students on both their progress and their writing skills.

We used a series of post-lecture prompts in a first year or foundation- level course. Students were invited to complete these (for a small credit) and text analysis of the results was used to direct the first portion of the following lecture. In another part of the project, students received writing prompts as part of their pre-lab work on Canvas in a large first year course. Just-in-time teaching (JiTT) has been posited to improve student preparation and motivation, as well as seen to promote ongoing formative assessment to allow instructors to target student weaknesses before the ‘too late’ of summative assessment. The project has showcased how automated text analysis can be used to facilitate JiTT approaches in large classes and has allowed students to access the benefits associated with the teaching approach. The student-directed feedback has allowed students to gain a more complex picture of their conceptual strengths and weaknesses than they would gain from traditional multiple choice questions.