Inductive analysis applied to the evaluation of a CAL tutorial
Atina Brooks, P. Vezza · Interacting with Computers · 1989
A human-computer interaction experiment is reported in which subjects used a CAL (computer-aided learning) tutorial to solve four problems about housing benefit regulations. The results are analysed inductively, with the analysis expedited by the use of inductive rule learning techniques. The results of the inductive analysis were unavailable at the time. Traditional statistics were used in the comparative evaluation of the CAL tutorial. Consequently a useful insight into the advantages of inductive analysis was provided and it is suggested that such an analysis might be of use for any HCI experiment. By way of a benchmark test, the result of applying inductive rule learning to Fisher's data is presented.