Stanton Amelia M, Meston Cindy M, Boyd Ryan L
Department of Psychology, The University of Texas at Austin , Austin, Texas.
Cyberpsychol Behav Soc Netw. 2017 Jun;20(6):382-388. doi: 10.1089/cyber.2016.0657. Epub 2017 Jun 1.
This is the first study to examine language use and sexual self-schemas in natural language data extracted from posts to a large online forum. Recently, two studies applied advanced text analysis techniques to examine differences in language use and sexual self-schemas between women with and without a history of childhood sexual abuse. The aim of the current study was to test the ecological validity of the differences in language use and sexual self-schema themes that emerged between these two groups of women in the laboratory. Archival natural language data were extracted from a social media website and analyzed using LIWC2015, a computerized text analysis program, and other word counting approaches. The differences in both language use and sexual self-schema themes that manifested in recent laboratory research were replicated and validated in the large online sample. To our knowledge, these results provide the first empirical examination of sexual cognitions as they occur in the real world. These results also suggest that natural language analysis of text extracted from social media sites may be a potentially viable precursor or alternative to laboratory measurement of sexual trauma phenomena, as well as clinical phenomena, more generally.
这是第一项研究，旨在检查从一个大型在线论坛的帖子中提取的自然语言数据中的语言使用情况和性自我图式。最近，两项研究应用先进的文本分析技术，研究有童年性虐待史和没有童年性虐待史的女性在语言使用和性自我图式方面的差异。本研究的目的是检验在实验室中这两组女性之间出现的语言使用差异和性自我图式主题的生态效度。从一个社交媒体网站提取存档的自然语言数据，并使用LIWC2015（一个计算机化文本分析程序）和其他单词计数方法进行分析。在大型在线样本中复制并验证了最近实验室研究中表现出的语言使用和性自我图式主题的差异。据我们所知，这些结果首次对现实世界中发生的性认知进行了实证检验。这些结果还表明，对从社交媒体网站提取的文本进行自然语言分析，可能是对性创伤现象以及更普遍的临床现象进行实验室测量的潜在可行的先导方法或替代方法。
