Abstract
To test the behavioral validity of the strategy method in a setting of operations management, we experimentally investigate competing newsvendor behavior under incomplete information with both the strategy method and the direct-response method. We observe that the ‘‘pull-to-center’’ effect exists only with low margin; mean order quantity with high margin does not significantly deviate from equilibrium prediction. We build a behavioral model based on overestimation and mean anchoring to explain competing newsvendor behavior. Estimates of the behavioral model confirm the existence of the behavioral biases. Meanwhile, order levels are not significantly different between the strategy method and the direct-response method. Hence, we suggest that the strategy method should lead to similar decisions in newsvendor settings compared to the direct-response method and may be adopted in most operations management settings associated to the newsvendor problem to improve the efficiency of experimental studies.
Acknowledgements
The authors thank the editors and anonymous referees for their comments and suggestions, which helped to improve the quality of the paper.