In recent studies, there has been focus on understanding learner performance and behaviour using Web-Based Instruction (WBI) systems which accommodate individual differences. Studies have investigated the performance of these differences individually such as gender, cognitive style and prior knowledge. In this article, the authors describe a case-study using a large student user base. They analysed the performance of combinations of individual differences to investigate how each investigated item influenced learning performance. The data was filtered to validate the data mining findings in order to investigate the sensitivity of the results. Moving data threshold was used to evaluate their findings and to understand what could affect the performance. The authors found that certain combinations of individual differences altered a learner's performance level significantly using Data mining techniques. They conclude that designers of WBI applications need to consider the combination of individual differences rather than considering them individually in measuring learners' performance.