Hypotheses testing
To test the hypotheses and the impact of controls on the dependent variables, a number
of multiple linear regression analyses were performed. First, we assessed the impact of
the control variables (demographics, tenure and nature) and employee attitudes and
behaviour. Models 1, 3, 5 and 7 (Table IV) show that age and gender have no significant
association with any of the attitudes or behaviour. Permanent work is negatively
associated with psychological empowerment and affective commitment, and tenure is
positively related to customer-oriented behaviour; however, the percentage of variance
explained by the controls cumulatively in all attitudes and behaviour was insignificant.