Abstract: Impression preference, the likes or dislikes when facing a stranger, is the assessment component of First Impression. The objective of this research is to explore the mental mechanism of the process in the view of relational self. If a stranger is similar with one’s Significant Other (SO), which will activate the corresponding relational self and, at the same time, the preference for SO will be transferred to the stranger. It’s transference, fundamentally an implicit and automatic process, and it’s worth to note whether it can be controlled or not. In this paper, the researchers planned to test and verify the above-mentioned process, especially its controllability. Based on the theory of relational self, 2×2×2 mixed experimental design was adopted with three independent variables: relativity (between-subject factor)，that is whom the stranger is relative to — SO or acquaintance; similarity (within-subject factor), that is whether the stranger is similar to SO/acquaintance or not; and hint (between-subject factor), that is whether to remind the similarity or not. One point should be noted that researchers designate mother as SO and classmate as acquaintance, the former suggests intimate relationship,while the latter just a nodding acquaintance. The dependent variable was impression preference for strangers. The two-session paradigm of transference study was used in the procedure. Firstly, in order to get participants’ initial data, information questionnaires about a specific SO/acquaintance were made based on literature and interview. Given that different role experiences could be a noise variable, then, 80 female participants were selected by clustering sampling from sophomore students. They filled in the questionnaires and wrote down their ID. After that, for each participant, a specific questionnaire of impression preference for strangers was made based on the information collected. Secondly, two weeks later, the participants took part in the posttest but they really didn’t realize its relevance to the pretest. This time they needed to determine their preference for two strangers respectively (the one was similar with the former SO/acquaintance, the other was not). Three questions were provided for each stranger, and scales ranging from 1 to 7, with higher values indicating more preference. The manipulation of “hint” variable was depended on the instruction. That is, before making decision, the participants were told to recall whether the stranger was similar to one of his SOs or acquaintances. It was expected that “hint” could interrupt the automatic mental process, so that participants could make more rational judgment. 　　　The conclusions drawn from the experiment: 　　　1. When the stranger was similar to one’s SO, compared to the dissimilar one, he would get more preference, F(1,76) = 22.463, p< .001,η2 = .228. On the contrary, whether the stranger was similar to one’s acquaintance or not, there was no significant difference in impression preference. 　　　2. Hint reduced the level of impression preference when the stranger was similar to SO. F(1,76) = 4.741, p< .05,η2 = .059. The research suggests that, SO could influence the impression preference for strangers, and the process could be controlled. There were two main contributions in this research. One was to reinforce the idea that SO could influence one’s transference to the similar strangers by adding the acquaintance factor; the other one was to suggest that the automatic process of impression preference could be controlled. In particular, the mediated effect of relational self was commented in the discussion section.

Key words: relational self, significant other, impression preference, transference

摘要： 以80名大学女生为被试，在关系自我理论的框架下，采用移情研究的双时段范式，以实验法考察了关系自我和重要他人对于陌生人印象偏好的影响。结果发现：在无提示的条件下，与重要他人相似的陌生人得到更多的印象偏好；在有提示的条件下，这种由相似性而产生的偏好会显著降低。证明了重要他人对陌生人印象偏好的影响，并讨论了关系自我在其中发挥的中介作用。

关键词: 关系自我, 重要他人, 印象偏好, 移情