Chen Stephanie Y, Ross Brian H, Murphy Gregory L
Department of Psychology, New York University.
Department of Psychology, University of Illinois at Urbana-Champaign.
J Exp Psychol Learn Mem Cogn. 2016 Jul;42(7):1050-67. doi: 10.1037/xlm0000222. Epub 2016 Jan 4.
Category information is used to predict properties of new category members. When categorization is uncertain, people often rely on only one, most likely category to make predictions. Yet studies of perception and action often conclude that people combine multiple sources of information near-optimally. We present a perception-action analog of category-based induction using eye movements as a measure of prediction. The categories were objects of different shapes that moved in various directions. Experiment 1 found that people integrated information across categories in predicting object motion. The results of Experiment 2 suggest that the integration of information found in Experiment 1 were not a result of explicit strategies. Experiment 3 tested the role of explicit categorization, finding that making a categorization judgment, even an uncertain one, stopped people from using multiple categories in our eye-movement task. Experiment 4 found that induction was indeed based on category-level predictions rather than associations between object properties and directions. (PsycINFO Database Record
类别信息用于预测新类别成员的属性。当分类不确定时，人们通常仅依赖最有可能的一个类别来进行预测。然而，感知与行动的研究常常得出结论，即人们近乎最优地整合了多种信息源。我们使用眼动作为预测指标，呈现了一种基于类别的归纳的感知 - 行动模拟。类别是沿不同方向移动的不同形状的物体。实验1发现，人们在预测物体运动时会跨类别整合信息。实验2的结果表明，实验1中发现的信息整合并非明确策略的结果。实验3测试了明确分类的作用，发现做出分类判断，即使是不确定的判断，也会使人们在我们的眼动任务中不再使用多个类别。实验4发现归纳确实基于类别层面的预测，而非物体属性与方向之间的关联。（《心理学文摘数据库记录》
