- 學位論文
利他懲罰與抗議訊息在最後通牒賽局中對決策的影響
The Effect of Altruistic Punishment and Protest Message
摘要
本研究的目的在探討利他懲罰以及抗議訊息的回饋對分配者在最後通牒賽局中分配自利程度的影響。實驗中共 169名受試者，分別被隨機安排至 2(有利他懲罰、無利他懲罰) x 2(有訊息回饋、無訊息回饋) 4種實驗情境中。每位受試者擔任分配者進行 2回合的最後通牒賽局遊戲，在他們提出第一回合的分配後，電腦會根據實驗情境給予利他懲罰與訊息回饋。根據第一回合分配比例的不同，我們將分配者分為自利分配者、公平分配者與利他分配者。研究假設無論是利他懲罰還是抗議訊息的回饋，皆會降低分配者再次進行分配時的自利程度。結果發現87位在第一次分配時較為自利的分配者，若遭到接受者的利他懲罰或是抗議訊息時，皆會顯著的降低在第二回合再次進行分配時的自利程度。兩者中，利他懲罰對於降低分配者在第二回合進行分配時自利程度的效果較抗議訊息回饋大，但是兩者並無交互作用。對於81位在第一回合分配時為公平或利他的分配者而言，接受者有沒有給予正面的訊息回饋並不會顯著影響他們再次進行分配時的自利程度。從分配者在進行分配時所花費的思考時間來看，公平分配可能是人們認知上的一種預設抉擇，所需的思考時間最短；而自利分配的決策者，可能是必須評估被拒絕的風險，平均所花費的思考時間最長，顯示人們自私的行為是一種經過計算過後有意的抉擇。
並列摘要
The purpose of this study was to investigate the effect of altruistic punishment and protest message on ultimatum games. 169 undergraduate and graduate students were randomly assigned to the 2 (altruistic punishment) x 2 (protest message) experimental conditions to work on two ultimatum games. All of participants played the role of the distribution of money in the game. After their decision in the first game, altruistic punishment (by rejecting the unfair allocation) and protest message were manipulated by the computer. Participants were classified into selfish, fair, generous classes based on their response in the second game. We expected that both altruistic punishment and protest message would compell the proposer to be less selfish in the second game. Data from the 87 selfish distributors showed that participants became less selfish (giving more) when their decision was rejected by the receiver in the first game. But majority of their distribution was still selfish. Protest message might decrease distributor’s selfishness, too. But the effect was smaller than altruistic punishment. For the participants who were fair or altruistic in first game. Receiving positive message did not significantly affect their allocation in the second game. Decision time data further suggested that being fair is the easiest decision to make whereas decision to be selfish or generous take longer time to make.