AN EXTENDED MODEL OF IS PROJECT CONTROL CECIL ENG HUANG CHUA Nanyang Business School Nanyang Technological University Singapore 639798 CHRISTINA SOH Nanyang Technological University HARMINDER SINGH Nanyang Technological University INTRODUCTION IS project controls research focuses on identifying the appropriate sets of controls for managing complex IS projects. The literature is dominated by the situational view of control, which argues that situational factors such as the visibility of outcomes or behaviors identify the appropriate controls to employ. The controls literature in management has two alternate views on control. The balancing view suggests that organizations must have a healthy mix of formal (i.e., behavior and outcome) and informal (i.e., clan and self) controls, while the more controls view argues that control predicts project performance. These alternate views can potentially shed light on unexplained IS project control phenomena. For example, because the balancing view advocates the need for balance (Cardinal, Sitkin, & Long, 2004), it explains why all control modes are present in every project- an explanation that is absent from the situational view. However, because each view suggests predictions and explanations that contradict other views, these alternate views cannot be simply integrated into IS project control. This research therefore has two objectives: 1. Demonstrate the usefulness of the two alternate views on IS Project Control 2. Reconcile the three separate views of control. The contribution of this research is a new indifference-curve based model of IS project control. Our model demonstrates that each view is a special case of the utility maximization behavior of an IS project manager. Ideas and concepts present in each view can be represented in our model. The model also provides additional insights to the problem of IS project control. PROJECT CONTROL Control theory addresses the methods by which principals can ensure that human agents perform assigned tasks appropriately (Eisenhardt, 1989). The principal is the manager or organization that wants to achieve some goal. The agent is the employee, or vendor tasked to achieve that goal. Control is required, because the agent’s desire to achieve the goal is substantively less than the principal’s. The agent thus only puts in required effort when controls are in place (Eisenhardt, 1989; Farrelly & Quester, 2003; Wright, Mukherji, & Kroll, 2001). Four distinct modes of control have been identified: (1) behavior, (2) outcome, (3) clan, and (4) self (Henderson & Lee, 1992; Kirsch, 1997; Ouchi, 1980). Behavior control refers to management of the work process. Rules and procedures are prescribed, and the agent is monitored to ensure that they are obeyed. In outcome control, the agent is rewarded or punished Academy of Management Best Conference Paper 2005 OCIS: E1