Reply and Extension.

The article presents information on questions raised by the discussant and conference participants. Some of the questions were: How does inherent risk relate to the control structure elements investigated? Would the significant differences among error rates per stratum at the account level similarly hold at the aggregate level as differences in control structure variables? Why do the findings diverge from inferences drawn in other research (e.g., Johnson 1987)? The findings suggest that public status is a relevant consideration in anticipating the control structure of a client, although such a status does not appear to track directly to different error rates.