A total of 300 questionnaires were distributed and only 80 usable responses were received.
As such the response rate is 27% which can be considered high for a mail survey. The
response rate of 27% is actually higher than the normal response rate of 10-20% using the
same method in Malaysia. As the respondents for this study are the decision makers or IT
personnel influencing decision making in investment in computerization projects they are
known to be less likely to respond to mailed questionnaires than people in the general
population (Hunt and Chonko 1987). Thus subject to the limitation of the low response rate,
the data can be useful to shed light on the issue that is important in this study.
The profile of the respondents and the organizations are presented in Table 3 and 4.
[Insert Table 3 about here]
[Insert Table 4 about here]
As suggested by Zou et al. (1997), owing to the lack of comparable data from the non-
responding firms, direct comparison of the responding and non-responding firms was not
possible. We used the wave analysis method with the Student’s t-test as the next best
approach to compare between the early and late replies as suggested by Armstrong and
Overton (1977). The wave analysis method assumes that those who respond less readily are
more like non-respondents. (Zou et al., 1997). They suggested using the t-test procedure
under the assumptions of both equal and unequal group variances. In the t-test analysis, we
found no between-group mean differences at the 5% level for any of the variables in the study.
Thus, it may be concluded that non-response bias was not of particular influence in this
research. (Skarmeas, Katsikeas and Schlegelmilch, 2002).