ISSN 0439-755X
CN 11-1911/B

心理学报 ›› 2023, Vol. 55 ›› Issue (11): 1780-1792.doi: 10.3724/SP.J.1041.2023.01780

• 研究报告 • 上一篇    下一篇


孙亚茹, 刘泽军, 段亚杰, 陈宁(), 刘伟()   

  1. 上海师范大学教育学院, 上海 200234
  • 收稿日期:2023-01-03 发布日期:2023-08-30 出版日期:2023-11-25
  • 通讯作者: 陈宁, E-mail:; 刘伟, E-mail:
  • 作者简介:孙亚茹和刘泽军为本文共同第一作者。

How collaboration reduces memory errors: A meta-analysis review

SUN Yaru, LIU Zejun, DUAN Yajie, CHEN Ning(), LIU Wei()   

  1. College of Education, Shanghai Normal University, Shanghai 200234, China
  • Received:2023-01-03 Online:2023-08-30 Published:2023-11-25


为探究协作记忆中错误修剪效应的稳定性和影响因素, 经文献检索和筛选, 对38项协作记忆研究的64个独立样本(总样本量N = 6225)进行元分析。结果发现, 协作记忆中的错误修剪和协作抑制均稳定出现; 调节效应分析表明, 协作方式能调节错误修剪, 但不影响协作抑制效应; 材料类型对错误修剪效应无显著影响, 但情景材料有更高水平的协作抑制; 熟悉关系增强错误修剪效应并削弱协作抑制。以上结果表明, 协作能稳定地抑制错误数量, 提升协作记忆的正确率, 但在一定程度上受到协作方式和关系类型等因素的调节。

关键词: 协作记忆, 元分析, 错误修剪, 协作抑制, 调节效应


In collaborative memory, the memory performances of collaborative and equal-sized nominal groups were measured by the number of correctly recalled items. By comparing the correct recall results between the two groups, collaboration during the retrieval phase is seen to possibly result in collaborative inhibition and collaborative facilitation. However, recall error items were also essential indicators of collaboration. Several studies have considered error recall items as indicators to show that collaboration is beneficial in reducing errors. The phenomenon of collaborative groups recording significantly fewer recall errors than nominal groups is referred to as the “error pruning effect.” The mechanisms and moderators of the collaborative inhibition effect have been explored in several previous studies, but evidence on the mechanism of the “error pruning effect” is scarce. This meta-analysis aimed to investigate the robustness of error pruning in collaborative memory and to examine the potential mechanisms and moderators.
Studies were identified with several keywords, including “collaborative memory”, “collaborative recall”, “collaborative inhibition”, and “collaborative facilitation”. English language databases, including Web of Science, Science Direct, EBSCO, and ProQuest, as well as the Chinese language database CNKI, were searched. From 38 empirical studies (from a total sample N = 6225), 64 independent samples were included. We chose the random-effect model to conduct the meta-analysis using CMA 3.3. The 64 independent samples showed considerable heterogeneity. Moreover, no substantial publication bias was found in the studies, which was confirmed by the funnel plot, fail-safe number, and trim and fill methods.
Standardized mean differences measured by Hedges’ g were used as the effect size index in the meta-analysis. The main effect showed a large and robust error pruning effect and collaborative inhibition effect in the results. Moreover, the results indicated that the collaborative inhibition effect commonly accompanies the error pruning effect. Further analysis revealed that collaborative approaches and interpersonal relationships moderate the error pruning effect. In particular, collaboration of free-flowing and consensus building enhanced the error pruning effect, while collaboration had no significant effect on the inhibition effect. The type of material had no significant effect on error pruning, while story material increased collaborative inhibition. Familiar relationships increased the error pruning effect, but they weakened collaborative inhibition.
Overall, the study results demonstrated the effect of collaborative recall on inhibiting error and improving accuracy. Collaboration and interpersonal relationships may act as important moderating variables in the process. Although error pruning resulted from a feeling of knowing through recall from collaborative partners, it required a relatively low level of processing. Lastly, efficient error correction could be easily achieved through sufficient communication.

Key words: collaborative memory, meta-analysis, error pruning, collaborative inhibition, moderating effect