Research method9

A controlled laboratory experiment was conducted to test the proposed hypotheses. The participants consisted of 122 undergraduate students enrolled in a management accounting course at a large Australian university. In a management accounting course, these students were taught about the concepts and issues relating to capital budgeting and various evaluation techniques.10 The average age of the subjects was 21.20 years. These subjects had, on average, 17.35 months of work experience. Seven participations responded incorrectly to one or both manipulation check questions, so their data were excluded, leaving 115 usable responses for final data analyses.

The decision scenario administered was adopted from Harrell and Harrison (1994). All participants assumed the role of a junior project manager with "The GlobalOne Company" and were asked to make a decision related to the continuation or discontinuation of a project that they had initiated and managed. The project, Project K, had a seven-year life span, and it was at the end of the fourth year that project managers were asked to make a decision. Information about the historical and expected future performance of Project K was provided. For instance, the project's expected net cash inflow had originally been $270,000 each year. During the past 4 years, the actual net cash inflow was $320,000, above the expected net cash inflow. However, due to a downturn in the economy, the junior project manager had already managed five failed projects. As for Project K, it was also predicted that for the remaining 3 years of its life, its expected net cash inflow would drop sharply to only $50,000 per year. The net present value (NPV) of the remaining three years' expected net cash inflow would be $144,327, while if this project were terminated, its salvage value in the remaining 3 years of its lifetime would be $177,500. Therefore the best decision from the firm's perspective would be to terminate Project K since its NPV was below its current salvage value.

In addition, apart from Project K's historical and future performance data, participants were also provided with information about GlobalOne's investment project policy, which they needed to consider in their decision-making process. The investment policy suggested that project managers should not discontinue more than five failing projects in any calendar year. The rationale for this policy was to discourage project managers from simply initiating new projects that might not be profitable for the company. Furthermore, unprofitable or failing projects that were terminated would be costly to the company. Any violation to this investment policy would attract severe disciplinary actions.11 The objective of this information was to elicit the presence or absence of obedience pressure.

Participants were randomly assigned to one of four treatments. These were generated by crossing two levels of information (public and private) with obedience pressure (absent and present). Participants assumed the role of a junior project manager with a growing reputation as a talented project manager. This had resulted in a competing firm, "NationalBest Company," initiating recruitment discussions for a better position with a substantially higher salary. If the unprofitability of a project were known publicly, it could damage the growing reputation and marketability of the junior manager and cause the competing firm to stop the recruitment discussion.

Participants in the public information condition were told that the information about Project K's unprofitable future performance was already known to others in the firm and industry, including the NationalBest Company. Participants in the private information condition were told that information about the project's unprofitable future economic viability was known only to them as project manager, and would not be known to others, including the NationalBest Company, until the project's completion in 3 years. Participants in the condition where obedience pressure was absent were told that the recommended investment policy was too restrictive and narrow minded; therefore this policy had been abolished. Instead, the chief executive officer (CEO) suggested that all project managers should make their own decisions independently based on all the relevant available information. Participants in the condition where obedience pressure existed were told that the CEO was very keen on the recommended policy, which had been implemented only 3 months ago. They were also told that if they chose to discontinue Project K, they were likely to face severe punishment resulting from their having violated the investment policy and disobeyed the CEO's command.

Participants were asked to respond to two manipulation check questions. First, they were asked whether the project's performance evaluation information was either known only to them as project manager, or widely known to others in their firm or industry. Second, they were asked whether they were experiencing pressure to obey the commands of their CEO and to conform to the company's investment policy.12 After completing these manipulation check questions, participants were also asked to complete the authoritarianism scale.

The dependent variable used in the experiment was the participants' preference to continue or discontinue an unprofitable project. The decision to continue or discontinue the project was indicated on a 10-point Likert-type scale numbered from 1 to 10. The scale 1 was labeled definitely continue, and 10 was labeled definitely discontinue. The three independent variables were (1) the availability of information (public or private), (2) obedience pressure (absent or present) and (3) the subject's personality trait of authoritarianism (high or low).