Analysing student responses to summative and formative assessments (Generic protocol G20-P3)

[Note: this protocol is a living document and the updated version can be sourced from the Loughborough Ethics Office]

The Mathematics Education Centre undertakes research which is focussed upon understanding and improving the learning of mathematics and statistics in the higher education sector. Key to this work are colleagues who research the design and impact of assessments used for summative and formative purposes in higher education. The Centre pioneers new approaches to assessing mathematics and statistics, with a particular emphasis on computer-aided assessment supported through the LEARN (Moodle) environment. For example, we have implemented the online STACK system for assessing students' mathematical knowledge, and this system is now widely in use across the university. Recently we developed a new Comparative Judgement system for peer assessment which is used to assess students' mathematical knowledge, and has the potential to be used for peer assessment across a wide range of higher education disciplines. We also have strong links with assessment innovations in mathematics and statistics at other universities in the UK. For example, colleagues in the Centre collaborate with colleagues at the University of Edinburgh to research both the STACK and Comparative Judgement systems. 

A rich source of evidence for such work is secondary data arising from students' responses to the assessment tasks that are routinely administered on modules across the university and at other universities. These tasks include, for example, online activities (e.g. STACK quizzes, Comparative Judgement tasks, Multiple-Choice tests) and paper-based activities (e.g. traditional exams, in-class tests, coursework).

This generic protocol is intended to cover studies in which researchers require either anonymised data (students' responses to the assessment tasks where identifying information, namely university ID numbers, has been first removed by Academic Registry) or pseudo-anonymised data (where identifying information, namely university ID numbers, is removed initial by the researcher, and in some cases where a codebook is used to track students across several assessments, but data will be anonymised in any outputs) that have been generated as part of normal university teaching and assessment practices. These data can include historic archives (e.g. the university stores sample exam scripts going back 6 years) and more recent student responses (e.g. a test taken recently); they can include student responses to summative assessments (i.e. that contribute to module scores) and formative assessments (i.e. that do not contribute to module scores). Specifically, this generic protocol is for those cases where no further data is required from students (e.g. demographic data, prior qualifications, interview data etc will not be collected), and so all data are secondary and anonymised or pseudo-anonymised, and therefore permission from students will not be sought.