Improving an online diagnostic test via item analysis

This poster reports on a project to apply the theoretical framework of psychometric testing to evaluate the performance of an online diagnostic test in mathematics, and to design and implement consequent improvements. The steps of the evaluation methodology are described, with a view to the techniques being more routinely used in the re-design of digital tasks. A key finding is that for the diagnostic test, an exploratory factor analysis gave results consistent with an entirely separate classification according to an established taxonomy.