MULTINOMIAL PROCESSING MODELS OF SOURCE MONITORING
MULTINOMIAL PROCESSING MODELS OF SOURCE MONITORING
复制标题
DOI：
10.1037/0033-295x.97.4.548
复制
发表时间：
1990-10-01
影响因子：
5.4
通讯作者：
RIEFER, DM
中科院分区：
文献类型：
--
作者：
BATCHELDER, WH;RIEFER, DM
本文提出了人类记忆中源监控范式的一系列处理模型。源监控和现实监控的特殊情况已成为非常流行的评估不同受试者群体记忆缺陷的范例。该范例提供了满足乘积多项式约束的分类数据，这非常适合使用处理树结构进行多项式建模，如 DM Riefer 和 WH Batchelder 中所述（参见记录 1988-31527-001）。本文开发的模型基于高阈值信号检测模型的思想，涉及项目检测参数、源识别参数以及反映猜测偏差的各种参数。这些模型的目的是提供单独的、基于理论的旧物品检测和来源辨别措施。这些模型可以加强基于临时统计的传统分析，并避免传统分析可能产生的有缺陷的解释。通过分析现实监控和双语记忆领域已发布的数据集，揭示了模型的实用性。（PsycINFO 数据库记录 (c) 2016 APA，保留所有权利）
This article presents a family of processing models for the source-monitoring paradigm in human memory. Source monitoring and the special case of reality monitoring have become very popular as paradigms to assess memory deficits in various subject populations. The paradigm provides categorical data that satisfy product-multinomial constraints, and this lends it nicely to multinomial modeling with processing-tree structures as described in DM Riefer and WH Batchelder (see record 1988-31527-001). The models developed herein are based on ideas from high-threshold signal-detection models, and they involve item-detection parameters, source-identification parameters, and various parameters reflecting guessing biases. The purpose of the models is to provide separate, theoretically based measures of old-item detection and source discrimination. The models may strengthen traditional analyses that are based on ad hoc statistics, as well as avoid flawed interpretations that the traditional analyses may produce. The usefulness of the models is revealed by analyzing published data sets from the areas of reality monitoring and bilingual memory.(PsycINFO Database Record (c) 2016 APA, all rights reserved)