Department of Psychology, Uppsala University, Sweden.
J Exp Psychol Learn Mem Cogn. 2010 Jan;36(1):1-16. doi: 10.1037/a0017893.
Although people often have to learn from environments with scarce and highly selective outcome feedback, the question of how nonfeedback trials are represented in memory and affect later performance has received little attention in models of learning and decision making. In this article, the authors use the generalized context model (Nosofsky, 1986) as a vehicle to test contrasting hypotheses about the coding of nonfeedback trials. Data across 3 experiments with selective decision-contingent and selective outcome-contingent feedback provide support for the hypothesis of constructivist coding (Elwin, Juslin, Olsson, & Enkvist, 2007), according to which the outcomes on nonfeedback trials are coded with the most likely outcome, as inferred by the individual. The relation to sampling-based approaches to judgment, and the adaptive significance of constructivist coding, are discussed. (PsycINFO Database Record (c) 2009 APA, all rights reserved).
虽然人们经常需要从缺乏和高度选择性的结果反馈的环境中学习，但在学习和决策模型中，关于非反馈试验如何在记忆中表示以及如何影响后续表现的问题，很少受到关注。在本文中，作者使用广义语境模型（Nosofsky，1986）作为一种手段，来测试关于非反馈试验编码的对比假设。在 3 项具有选择性决策相关和选择性结果相关反馈的实验中收集的数据，为建构主义编码假说提供了支持（Elwin、Juslin、Olsson 和 Enkvist，2007），根据该假说，个体推断出最有可能的结果来对非反馈试验的结果进行编码。本文还讨论了与基于抽样的判断方法的关系，以及建构主义编码的适应性意义。（《心理学文献摘要数据库》（PsycINFO Database Record），2009 年 APA，保留所有权利）。
