Abstract: Previous studies showed that self-related information will be processed elaborately to gain an advantage in memory performance. This phenomenon is called self-reference effect (SRE), and the effect also exists at the group level. The memory advantage of the in-group information is called the group-reference effect (GRE). Theoretically, self-construal is considered to be one of the variables that influence the in-group reference effect. However, the question of whether the self-construal influence the out-group reference effect is still open. Therefore, the current study focuses on the promoting effect of self-construal on memory performances of out-group information. Two experiments were administered to explore whether self-construal could promote out-group reference effect. Experiment 1 recruited Han college students (N=119) who were randomly divided into 3 self-construal conditions: independent, interdependent, and neutral. Four orienting tasks were arranged: self-referential processing, Han-referential processing, Naxi-referential processing and Luxun-referential processing. During each task, participants were asked to make judgment in a scale of 5 points on the current reference object using the adjective words displayed on computer screen one by one in a rate of 1 word/4 seconds. After that, participants will be required to judge whether the adjective words had been displayed. Experiment 1 found that the recall rate of self-referential was significantly higher than in/out-group-referential and other-referential in all conditions. At the same time, there is an interaction effect between self-construal and referential conditions. In the interdependent self-construal condition, the in-group reference will no longer have the memory advantage over the out-group referential. Experiment 2（N=183） used the retrieval-induced forgetting paradigm to repeatedly check the results of Experiment 1. A total number of 183 college students of ethnic Han took part in Experiment 2, which took a 3 (self-construal: interdependent self-construal vs. independent self-construal vs. neutral) x 2 (ethnic: Han vs. Naxi) design between participants. Experiment 2 includes two sections. In the first section, participants will be required to learn the typical name about Han and Naxi ethnic to ensure that typical name can be regarded as representation of ethnic by participant. There are four phases in the second section. (1) Study phase: the computer will display the Chinese characters, The 32 exemplar divided into two categories and were instructed to memorize by making sentences with 2 different referential conditions ; (2) Retrieval-practice phase: One of the category will be selected randomly in this phase, and half of those exemplar will be practiced randomly also . Each cue comprised a category name and the first initial character of an exemplar and the participants were asked for written recall of the target exemplar in response to each cue; (3) Distraction phase: The participants were asked to perform random arithmetic within 3 minutes; (4) Final test phase: The participants were required to retrieve written recall as exemplars as much as possible in response to each presented category name. The results found that the RIF was not observed under the outgroup-referential under the interdependent self-construal condition, which means there is a significant out group-reference effect when the interdependent self-construal is prime. The two experiments consistently showed that interdependent self-construal can make the out-group information be elaborately processed, so that there is the same memory performance in the reference information of the in-group and out-group.

Key words: self-construal, in-group reference effect, ut-group reference effect, retrieval-induced forgetting

摘要： 内群体身份被视为群体自我，也存在与个体自我类似的信息精细加工效应，并受到文化的影响。自我建构是文化在个体层面的体现，本研究通过两个实验考察自我建构内、外群体参照信息加工的作用。实验1选用记忆参照实验，实验2选用提取诱发遗忘实验，通过代词圈点法启动自我建构，两个实验共同发现：独立自我建构启动时，存在内群参照效应；而互依自我建构启动条件下，内群体信息加工优势消失，没有出现内群参照效应。

关键词: 自我建构, 内群参照效应, 外群参照效应, 提取诱发遗忘