以期望理論分析愛荷華賭博作業之檢定力

Abstract

愛荷華賭博作業是一個廣泛用來偵測可能決策缺失的認知作業。透過以期望理論對愛荷華賭博作業進行認知模擬，期望理論的參數分別對應了愛荷華賭博作業心理歷程的不同影響成分。當比較兩個群體在愛荷華賭博作業上的歷程時，典型的分析方式是二階段的。第一階段先以個別受試為單位估計參數，第二階段則對此參數進行平均數的比較。鄭中平（2007）修正了原始的期望模型，將個別受試參數視為由一分配中抽出，亦即，考慮參數的隨機效果，因此不同群體參數平均數的比較可以在一階段完成。由於不同群體的決策歷程差異可能是許多心理學研究興趣所在，偵測群體差異的檢定力議題不容忽視。本計畫目的在以蒙地卡羅法比較兩階段與一階段分析的檢定力。

Iowa Gambling Task is a cognitive task designed for exploring the possible decision making deficit. By modeling this task, parameters of the expectancy-valence model are correspondent the components of psychological processes underlying the Iowa gambling task. When comparing psychological process of different groups of participants in the Iowa Gambling Task, the typical analysis is two-staged. The first stage is to estimate the parameters individually. The second stage is to compare the parameters from different groups. Cheng (2007) modified the original Expectancy-Valence model such that parameters of different participants in the same group are considered to be normal distributed. By doing so, comparison of parameters from participants of different groups can be made in one stage. For the difference of decision making processes of different populations may be the main concern of many psychological studies, power to detect the difference cannot be overlooked. The aim of the study is to compare the powers of two-stage and one-stage analyses by Monte Carlo method.