Method
Six different scales were used to capture the variables of interest in this study:
individualism-collectivism, work-pace, Kiasu-positive, Kiasu-negative,
task-participation and social-support. The individualism-collectivism measure
was based on theWagner (1995) preference for group work subscale; work-pace
was based on the work-pace subscale of Schriber and Gutek (1987);
Kiasu-positive and Kiasu-negative were adapted and modified from Hwang
et al. (2002) study, while the task-participation and social-support subscales
were based on items by Campion et al. (1993), Ibbetson and Newell (1996) and
Watson et al. (1991) after examination for similarities and differences and
consequent elimination of overlaps. To ensure that respondents who were not
conversant with the English language could participate in the study, the scales
were translated into Chinese. Back-translation was done to minimize differences
of scales in the two languages. All items were anchored on a seven-point Likert
scale from 1 (strongly disagree) to 7 (strongly agree). All the scale items were
first subjected to a reliability (Cronbach Alpha) test and also an initial Lisrel
measurement model fit test. Less reliable items were eliminated until there were
at least three items to each construct. While the reliability coefficients for these
scales were moderate – ranging from 0.69 to 0.77, Lisrel measurement model fit
indices (Joreskorg and Sorbom, 1989) for a fully constrained measurement
model indicated acceptable fit (Chi-square (759.54)/df (230) ¼ 3.30; GFI ¼ 0.84;
AGFI ¼ 0.81; RMR ¼ 0.11) for subsequent structural model development and
hypothesis testing. Scale item descriptions, means, standard deviations and
Cronbach Alpha coefficients are presented in the Appendix