Department of Computer Science in Economics, Faculty of Economics and Sociology, Institute of Logistics and Informatics, University of Lodz, Lodz, Poland.
PLoS One. 2020 Feb 25;15(2):e0229354. doi: 10.1371/journal.pone.0229354. eCollection 2020.
Systematic exposure to social media causes social comparisons, especially among women who compare their image to others; they are particularly vulnerable to mood decrease, self-objectification, body concerns, and lower perception of themselves. This study first investigates the possible links between life satisfaction, self-esteem, anxiety, depression, and the intensity of Instagram use with a social comparison model. In the study, 974 women age 18-49 who were Instagram users voluntarily participated, completing a questionnaire. The results suggest associations between the analyzed psychological data and social comparison types. Then, artificial neural networks models were implemented to predict the type of such comparison (positive, negative, equal) based on the aforementioned psychological traits. The models were able to properly predict between 71% and 82% of cases. As human behavior analysis has been a subject of study in various fields of science, this paper contributes towards understanding the role of artificial intelligence methods for analyzing behavioral data in psychology.
系统接触社交媒体会导致社会比较,尤其是在女性中,她们会将自己的形象与他人进行比较;她们特别容易情绪低落、自我客体化、关注身体问题,并对自己的评价降低。本研究首先通过社会比较模型,调查生活满意度、自尊、焦虑、抑郁与 Instagram 使用强度之间可能存在的联系。在这项研究中,974 名年龄在 18 至 49 岁之间的 Instagram 用户自愿参与,完成了一份问卷。结果表明,分析后的心理数据与社会比较类型之间存在关联。然后,实施了人工神经网络模型,根据上述心理特征来预测此类比较的类型(积极、消极、平等)。这些模型能够正确地预测 71%到 82%的案例。由于人类行为分析已经成为各个科学领域的研究课题,本文为理解人工智能方法在心理学中分析行为数据的作用做出了贡献。