Adon Moskal is helping lecturers refine their teaching through analysis of student answers.

In large classes especially, it is difficult for lecturers to get a good sense of how well students understand the concepts they are being taught. If they could analyse the information which students provide in responses to short answer questions (about 50 to 100 words), that would inform lecturers whether and which concepts needed to be explained more clearly, to help them improve their teaching.

So Adon Moskal, a lecturer in Information Technology, with Jenny McDonald at Auckland University, has developed software that provides text analytics of students' answers. The software can provide lecturers with top frequencies for single words, pairs of words, and trigrams. If those are not consistent with what the lecturer expected, they are able to consider why that might be so and address that in future classes with the students. If a lecturer also uploads their teaching materials - notes and set reading for example - the software will do a comparison to find out where the highly frequent word combinations are found in the teaching materials. This could reveal if students have misunderstood what a lecturer was meaning.

The software is useful not just for analysis of student learning, but also of student evaluation data. Identifying high frequency themes in student responses to evaluation of their experiences can help reveal issues that might need to be addressed to remove barriers for students and help them succeed.

The software is open source, via the Quantext website. It is currently being piloted at universities in New Zealand and internationally. Work is well under way on a new version with an improved interface based on lecturer feedback about useful functionality. 

July 2018