社交媒体中人格特质推理的准确性、过程及影响

The ability to infer personality traits is a highly evolved aspect of human cognition. Researchers are increasingly interested in exploring the capability of personality trait inferences on the basis of social media with its wide application. Past research investigated whether or not individuals were able to infer some personality traits of strangers after observing their information from social networking; the present project focuses on how well individuals infer the low, middle and high levels of a given personality trait based on limited information of Wechat. Our previous studies revealed a ‘U’ pattern of accuracy in personality trait inferences, in which perceivers were effective in inferring the low or high levels of empathic trait and the big-five personality traits but were unable to infer the middle after observing a brief sample of behavior..Based on these backgrounds, the present project will adopt behavioral experiments combining with the eye-tracking technique to investigate the psychological mechanisms of personality trait inferences from social media. Results of the pilot study replicated the ‘U’ pattern, demonstrating that perceivers were accurate in inferring the low and high empathic trait but inaccurate in inferring the middle. Study 1 will explore the scope and the limitation of the capability to judge strangers on trait continua. On the basis of the social relation lens model, Study 2 will pinpoint the influences of the ‘U’ pattern from the process of trait inferences from a thin slice of information of Wechat; Study 3 will investigate the social consequences of personality trait inferences. The personality trait judgment task is used in each study for measuring the ability to make trait inferences. Accuracy is defined as the self-other agreement of the personality measurement. .This project is important for its theoretical and practical significance. It will help to build up testable models of trait inferences and serve to guide our everyday practices of trait judgments on social media. In addition, it will provide scientific bases for achieving the goal of ‘strengthening the construction of Internet contents’.

人格特质推理是一种重要的社会认知能力。随着社交媒体的广泛应用，研究者开始探讨个体从社交媒体信息能否推测陌生人的人格特质；本项目则关注个体从有限的微信信息能否推测人格特质的低、中或高水平。申请人的研究发现，个体在现实情境中能够推测共情和大五人格特质的低和高水平，但不能推测中间水平，即特质推理能力呈现‘U’型模式。.在此基础上，本项目拟采用行为实验与眼动追踪技术相结合的方法，揭示社交媒体中人格特质推理的机制。预实验的结果表明特质推理‘U’型模式泛化到社交媒体情境。研究1揭示这种能力的广度和精度；研究2以社会关系透镜模型为依据，从特质推理的过程揭示为什么呈现‘U’型模式；研究3考察该模式对社交生活的影响。研究通过特质判断任务测试特质推理的准确性，以自我-他人报告一致性作为准确性指标。.本项目将有助于完善人际知觉模型的可预测性，改善人们在社交媒体中的交往，为实现‘加强互联网内容建设’提供科学依据。

本课题基于人际关系透镜模型以及人格特质推测的现实准确性模型，探讨在社交媒介（微信）情境中，人们对人格特质不同水平的推测准确性、推测过程、以及人格特质第一印象对人际交往的影响。研究1探究通过观察陌生人微信的基本个人信息，人们在多大程度上正确推测其大五人格特质维度的低、中、高水平。研究2中，互不相识的双方分别在微信信息聊天及面对面交谈情境中推测对方的人格特质，以探究人格特质第一印象在社交媒介与现实情境中的一致性，并且考察人格特质推测过程个体对可观察线索的利用。研究3考察基于微信信息对他人形成的人格特质第一印象在多大程度上能够预测人们随后与陌生人进行社交互动的意愿。研究的结果表明：（1）即便通过观察陌生人非常少的微信信息，人们能够对其大五人格某些特质的低或高水平形成准确的第一印象，而普遍不能正确推测人格特质的中间水平；（2）在微信信息互动及观察对方与他人面对面交谈的情境中，人们能够对陌生人的大五人格及共情特质形成一致的印象，并且，每种情境中某些可观察的线索能够预测人格特质的第一印象；（3）观察陌生人微信有限的信息，人们对热情-能力特质形成积极的第一印象，并且这种印象在一定程度上预测他们随后与这些陌生人交友或合作的互动意愿。本课题通过一系列研究，从人格特质推测的准确性、形成过程及影响等方面探讨了基于社交媒介的人格特质第一印象。这些研究丰富了社会知觉理论，将现实情境中对不同人格特质水平推测准确性的U型模式拓展到社交媒介环境，并在网络社交情境检验人际关系透镜模型及现实准确性模型的效力。此外，本课题为基于社交媒介研究人格特质第一印象提供了新的视角及可行的研究方法，从人格特质不同水平、虚拟与现实情境人格特质第一印象的一致性、第一印象导致的社交后果等方面开展实验研究。这些研究探索不仅为理解人类社会知觉能力的广度和精细度提供了新论据，而且对人格特质第一印象领域未来的研究具有一定的启发。