ISSN 0439-755X
CN 11-1911/B

心理学报 ›› 2015, Vol. 47 ›› Issue (1): 19-28.doi: 10.3724/SP.J.1041.2015.00019

• 论文 • 上一篇    下一篇



  1. (1北京师范大学发展心理研究所, 北京 100875) (2江西科技师范大学教育学院, 江西 330013) (3香港中文大学心理学系, 香港)
  • 收稿日期:2014-03-26 出版日期:2015-01-26 发布日期:2015-01-26
  • 通讯作者: 王大华, E-mail:
  • 基金资助:


Age Alters the Effects of Emotional Valence on False Memory: Using the Simplified Conjoint Recognition Paradigm

XIAO Hongrui1; HUANG Yifan2; GONG Xianmin3; WANG Dahua1   

  1. (1 Institute of Developmental Psychology, Beijing Normal University, Beijing 100875, China) (2 School of Education, Jiangxi Science & Technology Normal University, Jiangxi 330013, China) (3 Department of Psychology, The Chinese University of Hong Kong, Hong Kong, China)
  • Received:2014-03-26 Online:2015-01-26 Published:2015-01-26
  • Contact: WANG Dahua, E-mail:


以康奈尔情绪DRM词表为主要实验材料, 使用简化的联合再认范式(DRM范式的变式), 结合多项式加工树建模的统计手段, 对34名年轻人(23 ± 2岁)和28名老年人(68 ± 5岁)的再认记忆进行比较, 以考察不同年龄组中情绪效价对错误记忆的作用机制。结果发现材料的情绪效价对错误记忆的影响存在显著年龄差异:(1) 积极情绪能够有效降低老年人的错误记忆, 其作用机制为积极情绪增强了老年人的字面痕迹而减弱了要点痕迹提取; (2) 消极情绪能够有效地降低年轻人的错误记忆, 其作用机制为消极情绪使年轻人反应偏向降低, 但不影响其记忆痕迹的提取。结果表明, 简化的联合再认范式下, 情绪效价对错误记忆的效应存在明显的年龄差异:老年人表现出积极情绪降低错误记忆的积极偏向; 年轻人表现出消极情绪降低错误记忆的消极偏向; 情绪效应对错误记忆的认知机制存在年龄差异。

关键词: 错误记忆, 情绪效价, 年龄差异, 简化的联合再认范式, 多项式加工树建模


Previous research has revealed robust differences between young and older adults’ accurate emotional memory. Typically, negative emotion prompts memory performance in young adults, while positive emotion benefits memory performance in older adults. Relatively, much less efforts have been devoted to investigating how emotion divergently influences false memory in older adults compared to young adults. The main purpose of the present study is to explore, first, how age alters the effects of emotional valence on false memory; and second, what are the cognitive mechanisms for the interaction effect of age and emotional valence on false memory. The Fuzzy Trace Theory (FTT) uses verbatim memory trace and gist memory trace to interpret individuals’ false memory, especially false memory under the framework of Deese-Roediger-McDermott (DRM) paradigm. According to recent research, we proposed some flaws of traditional paradigms (such as the Remember-Know paradigm) in the area of false memory; and we also argued that it would be critical to include response bias in addition to verbatim and gist memory trace to investigate false memory. To fulfill this need, the simplified conjoint recognition (SCR) paradigm, combined with a statistical method of multinomial processing tree model, was used in the present study to investigate false memory and its associated cognitive mechanisms. A sample of 34 young adults (aged 23 ± 2 years) and 28 older adults (aged 68 ± 5 years) completed the SCR task. In the task, the Cornell/Cortland emotion word lists, along with neutral word lists adopted from previous research, were implemented as experimental materials. The results displayed a significant interaction effect of emotional valence and age on false memory (i.e. false alarm). Specifically, positive emotion decreased older adults’ false memory, while negative emotion decreased young adults’ false memory. The method of multinomial processing tree model was further employed to model, parameterize and inference the cognitive mechanisms related to the effect of emotional valence on false memory within young and older group. It was found that negative emotion contributed to a lower level of false memory in young adults by lowering their response bias toward negative words. As to older adults, positive emotion boosted retrieval of verbatim memory trace and hampered retrieval of gist memory trace, then led to less false memory for positive words. Our study uncovered interesting age-related differences in emotional false memory; it also confirmed the necessity of a simultaneous consideration of verbatim memory trace, gist memory trace and response bias when investigating false memory.

Key words: false memory, emotional valence, age-related difference, simplified conjoint recognition paradigm, multinomial processing tree model