社會科學院: 經濟學研究所指導教授: 王道一; 馮勃翰謝富文Hsieh, Fu-WenFu-WenHsieh2017-03-032018-06-282017-03-032018-06-282015http://ntur.lib.ntu.edu.tw//handle/246246/275260我們重現王道一等人(2010)設計的傳訊賽局實驗，讓一個知道真實狀態的傳訊者送訊息給（試圖按照真實狀態）做決定的接收者，但兩人之間有利益衝突：傳訊者有誘因吹噓真實狀態。我們的實驗結果顯示台灣受試者的行為模式與文獻類似，同樣都傾向「過度溝通」──所傳訊息比均衡預測揭露更多關於真實狀態的資訊。我們也能根據level-k模型把每個受試者分類為不同思考層次的類型。我們在做完重現實驗之後緊接著另外做只有三種真實狀態的簡化版實驗（並加上兩種吸引受試者注意力的特異回合）。我們發現簡化版實驗的結果比較接近理論預測，而且會有更多的受試者被我們歸類成可以想兩層的類型L2。We replicate the experiment designed by Wang, Spezio, and Camerer (2010), in which an informative sender advises an uninformed receiver to take an action (to match the true state), but has incentives to exaggerate. We find similar behavior patterns with Taiwanese subjects. In particular, we also find “over-communication”—messages reveal more information about the true state than what equilibrium predicts, and classify subjects into various level-k types. In addition, we conduct a Simplified version of the same experiment with only three states (and two sets of “catch” trials to keep subjects attentive). We find the results are more close to prediction of equilibrium model in this simplified replication, and more senders are classified as L2.978416 bytesapplication/pdf論文公開時間: 2015/3/13論文使用權限: 同意無償授權傳訊者-接收者賽局策略性訊息傳遞說謊測謊實驗室實驗Sender-Receiver GameStrategic Information TransmissionLyingLie DetectionLaboratory Experimentthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/275260/1/ntu-104-R01323046-1.pdf