本研究針對決革嵌陷行為的成因進行分析，探討決策者在面對一連串負面結果，為什麼仍然執著於先前的抉擇，投入更多的資源與心力，而導致越陷越深的理由。就理論而言，Staw的申辯理論採動機情緒的觀點，認為「自我申辯」是促成嵌陷行為發生的主要機制；Kahneman 典Tversky 的前景理論則採訊息處理的觀點，主張「確定效果」的作用會引發知覺價值扭曲，這才是導致在嵌陷行為發 生的主因。本研究嘗試統合此二理論的想法，來解釋嵌陷行為的發生。研究採實驗室研究的方式，以中級以上主管為受試對象，並且用電腦模擬組織決策情境， 操弄個人責任的高低與實行成果的成敗兩變項，結果發現，企業組織中管理者的決策確實有嵌陷行為的發生，但是只有前景理論的「確定效果」獲得證實。申辯理論中的「自我申辯」效果並不被支持。最後，研究者並提出對缺陷行為的預防措施，以及發生嵌陷行為之後的修正之道。

"Decision entrapment" refers to the phenomenon that decision-makers commit to the negative or failure choices, and tend to throw more resources in some projects that seem to fail certainly. We try to combine two contradictory viewpoints, self-justification vs. information-processing perspectives, to clarify the causality of decision entrapment through experimental approach. Staw suggests that the emotional “self-justification" is the major mechanism of decision entrapment. Adopting the information-processing perspective, Kahneman & Tversky propose that “certainty effect" bring about perception distortion, which is the main cause of entrapment behavior. We used computer programs to simulate a decision-making scenario, and manipulated “personal responsibility" and "project output" variables. The subjects are middle-leveled (and above) managers from 5 companies, and were asked to play the role of general managers in the simulated task. The result doesn't support the self-justification perspective, but verifies certainty effect of prospect theory. We then propose some management techniques to prevent and correct entrapment behavior.