Kimura, K. （木村邦博） (2020).
Japanese Women’s Attitudes Toward Childrearing: Text Analysis and Multidimensional Scaling.
子育てに対する日本女性の態度：テキスト分析と多次元尺度構成法
Advanced Studies in Behaviormetrics and Data Science: Essays in Honor of Akinori Okada, 423-438.
https://doi.org/10.1007/978-981-15-2700-5_26
This study aimed to understand Japanese women’s attitudes toward childrearing via text analysis. The text data were taken from the emails that readers had submitted to a magazine as a reply to the call for essays on the theme “Is Childrearing a Strain?” I hypothesized that Japanese women tended to attribute “failure” to external factors such as institutional flaws and “success” to internal factors such as their personal conditions. I employed an “appearance of strings” approach to natural language processing to map the keywords onto “topics.” Based on the multiple classification of emails with respect to the topics referred, I calculated symmetric and asymmetric versions of Jaccard Similarity Coefficient and applied four multidimensional scaling models: Torgerson’s method, SMACOF, slide vector model, and drift vector model. There was a contrast between positive and negative feelings. There were also personal, interpersonal, and societal facets. The emails expressing positive attitudes tended to refer to only personal topics, while those manifesting negative attitudes tended to refer to the topics in all the three facets. In the latter group of emails, reference to personal topics implied reference to interpersonal topics, and reference to interpersonal topics implied reference to societal topics.