Table IV shows the results of the regression for all respondents. In this broader analysis, we were especially interested in the overall difference between respondents in public accounting and respondents in non public accounting settings.
Hence, the independent variable, public practice organization, was used but not the accounting firm type variables, such as Big Four accounting firm.
Also, client commitment, applicable only for respondents in public practice, could not be included (as it would have removed all non accounting firm respondents from the listwise deletion method regression).
In this regression, we also investigated the effect of the interaction term between public practice organization and proportion of time spent in current job in accounting/ auditing (re: H1d).
The coefficients for both the main effects as well as the interaction effect for public practice organization and proportion of time spent in current job in accounting/ auditing are all significant.
When there is a significant interaction effect, the main effects in the overall regression cannot be explained independently.
In this case, the interaction effect can be taken to mean that the effect of proportion of time spent in current job in accounting/auditing on independence commitment is conditional on the organizational type.
To more thoroughly analyze the interaction effect, it would be helpful to take a look at a general regression equation with one interaction term: where y = dependent variable, b’s = regression
coefficients, x’s=independent variables.
For this case, we let x1 represent public practice organization and x2 represent proportion of time spent in current job in accounting/auditing.
We can come up with an equation for non-public practice firms (with x1=0, meaning the interaction term is 0 as well),