LIMITATIONS OF THE STUDY
Cher Ping cautions that care should be taken in interpreting the results of his research. First, he points out that the sample size is very small and that might have affected both the reliability and validity of the research. However, as we have seen, in order to ensure more reliable results, the analysis was submitted to a t-test with a small p value. The level of significance was kept low at 1 per cent which gave 99 per cent probability that the results had not arisen by chance.
Second, he points out that there was always the possibility that improvement in academic results might have been due to the novelty effect which might soon wear off.
Third, only students from four Science classes in the college participated in the study and it would be difficult to generalize or even relate the research results to Arts and Commerce students who might experience a higher level of computer anxiety.
Fourth, the research only suggested that CBL in support classes had a positive impact on students' cognitive and attitudinal gains. It overlooked the teacher's attitudes towards the use of CBL as a teaching tool. The fact that certain pedagogical features appeal to students is no guarantee that a particular technique will be attractive to teachers because of the demands made on teaching time. Additional research would be needed before educators and