The Automated Detection of Gender Bias Patterns in Children's Books and Stories

Abstract

From the late 1960s through the 1970s, researchers worldwide have shown interest in the exploration of gender representation in children’s literature, including books, stories, and educational materials. A significant representational discrepancy was witnessed and proved between both genders in central characters, illustrations, titles, and text in different children’s stories and books through several studies conducted over the years. Several methods have been used for the detection of gender bias, yet most of these methods followed a manual frequency-based qualitative and quantitative content analysis approach that focuses on the word-level detection of gender bias in language. This study, however, presents an advanced automated computer-driven approach that can detect different gender bias categories at a phrase-level and sentence-level. This study applies its automated methodology and finds countless instances of gender bias patterns investigated in more than 200 children’s books and stories, most of which are still read to and by children today. It also tries to explore any relationship existing between the gender bias categories detected and some attributes collected, such as “author’s gender”, “book genre”, and “year of publication”. This study finds significant effects of the” author’s gender” and “book genre” on the use of the different types of gender bias categories where male authors tend to display a greater bias in language towards males as compared to female authors. This research also presents the previous work that has been done in the field of gender research in children’s literature and discusses the negative impact that a gendered language has at a micro-level and macro-level. Finally, this work aims to enhance the existing detection approaches, especially for the identification of gender bias existing at the level of the language, and it presents an automated machine-led content analysis approach for this purpose.