The results reported in Table 6 indicate that student prior knowledge and international status are significant determinants of MarlinaLS usage, while there was a weak association between gender and MarlinaLS usage. In interpreting these results, two points are worth noting. First, to the extent that the student result in the prerequisite subject Accounting 1B may also proxy for other intangible attributes such as diligence, commitment and the perceptions of students of the usefulness of on-line learning systems, the significant coefficient on PRIOR1B is to be expected. Second, there is little indication in Tables 3–5 that either female or international students demonstrating greater usage of MarlinaLS performed significantly better, on average, on the semester exam than did students within these categories using the system to a lesser extent. Accordingly, we conclude that the primary results of the study do not appear to be driven by these
sub-groups within the student cohort.