Content area
الخلاصة
A between subjects experimental design was constructed to measure decision makers' reinvestment behavior in a sunk cost decision relative to an alternative choice and a reserve fund choice. The independent variables were responsibility, decision choice relative-risk, and decision return. The dependent variable was the amount reinvested to a sunk cost decision. Second order interactions of responsibility, relative-risk, and return were hypothesized.
Alternative relative-risks and expected returns were manipulated by responsibility to reflect the kinds of decision alternatives strategic decision makers face. As such, opportunity costs were built into the study, contrary to conventional escalate-or-withdraw designs.
Planned comparisons of responsibility/relative-risk reinvestment means were conducted at levels of decision return. These tests indicated that responsible and nonresponsible subjects alike took a conservative reinvestment posture when they faced a risky reinvestment versus a conservative alternative choice in a low return state. In a high return state, responsible subjects escalated commitment to the sunk cost when they compared both decisions as low risk versus both decisions as high risk. In a high return state, nonresponsible subjects de-escalated when making the same comparison, however.
In a low return state the comparison of low reinvestment risk/high alternative risk versus high reinvestment risk/low alternative risk decision choices yielded significant de-escalation by both responsible and nonresponsible subjects.
These interactive differences indicate that the collective impacts of responsibility and relative-risk are influenced by level of return. An interaction of responsibility, relative-risk, and return was therefore demonstrated by planned comparisons.
These results contradict prospect theory predictions which maintain that decision makers will be risk-seeking in the face of negative feedback. Contrarily, subjects predominantly took a conservative posture when they analyzed marginal gains and losses from the perspective of partial loss recovery or when they analyzed marginal gains and losses from the perspective of complete recovery with potential gains. These results suggest the future utility of approaching the escalation phenomenon from the perspective of risk-return perception.
