To see if overall independence commitment is significantly lower for public practice firms (having controlled for other variables), again, we will look at both the main effect and the interaction effect. Regression equations similar to those of the above can be set for different values of proportion of time spent in current job in accounting/auditing. When the proportion of time spent in current job in accounting/ auditing equals zero, only the negative significant main effect coefficient needs to be interpreted, and in this case, it means that the predicted indepen- dence commitment value for public practice organizations is 0.8 units (out of a maximum of 20) lower than that for the non-public practice
organizations. With the interaction effect also being negative and significant, one can infer that as the proportion of time spent in current job in accounting auditing increases, the overall negative effect will be made even more pronounced. Hence, the results support H1a. In other words, for any level of pro- portion of time spent in current job in accounting/audit- ing, the slope of the regression line (independence commitment against public practice organization) remains negative.