Using computer assisted assessment to facilitate resource selection

The number of computer mediated learning resources is growing at a phenomenal rate, and the advent of the web has made access easier than ever. This has however led to difficulties identifying resources suitable for the particular teaching purpose needed. One way round this is through directories, however the information that they provide is often limited to a few parameters. Where human evaluation is included, the review is often too short to give meaningful advice on its quality and most appropriate usage, or too long to enable quick identification. Modern computer assisted assessment packages are capable of storing and analysing vast amounts of information on student learning. With appropriate analysis this data can be used to pinpoint the strengths and weaknesses of individual students and match these to learning resources that meet their needs. This paper outlines a proposal to use the analysis of Rasch residuals to compile profiles of individual students. Rasch analysis allows computation of a candidate’s average ability and a question’s average difficulty. Where an individual candidate’s ability on one question is below their average, they can be considered to be weaker on that question than on others. Students can then be directed to resources that address that weakness. This overcomes the problems of manually identifying student needs. Computer assisted assessment generates a great deal of data which can be utilised far more fully than is currently the case. This paper urges the CAA community to look beyond the creation of items and towards their future analysis.