Unveiling gender disparities: a visual representation of middle school science textbooks in China and the United States

• In Event: Gender Justice in Educational Access and Equity in Global Perspective

Proposal

The issue of gender representation in science education, despite decades of discussion and research, persists with a sense of urgency (Scantlebury 2014; Stoet and Geary 2018; Morales et al. 2023). Despite the increase in women’s enrollment in universities from 46 percent in 1985 to 56 percent in 2017 (González-Pérez et al. 2020), achieving gender parity in STEM (science, technology, engineering, and mathematics) remains a significant challenge. This challenge is particularly critical in the context of the growing demand for digital technologies, where the underrepresentation of women exacerbates existing gender divides and limits opportunities for innovation (ITU 2024). To meet the increasing demand for STEM professionals, it is essential to actively engage more girls and women in these fields from an early age (UNICEF 2023). This lack of parity often begins at the middle school level, indicating early educational influences (Jean et al. 2015; UNESCO 2015; McGee and Bentley 2017; Yatskiv 2017; Mahon and Murphy 2019).
Although both China and the United States (U.S.) have implemented significant reforms in STEM education—China with the “National Curriculum Reform Basic Program (Trial)” in 2001, a program aimed at modernizing and standardizing the national curriculum, and the U.S. through efforts by the National Research Council, an organization that provides leadership in education and research—these changes have not yet fully bridged the gender gap in STEM (Ministry of Education of the People’s Republic of China 2001). Despite increased educational and career opportunities for women, girls’ involvement in these fields remains limited due to societal gender biases, a shortage of female role models, and girls undervaluing their own skills (Msambwa et al. 2024). In addition, while students in the U.S. may demonstrate greater overall engagement with STEM fields, the gender disparity in aspirations is significantly more pronounced compared to China (OECD 2018a,b,c). Such a divergence indicates cultural, educational, and possibly systemic factors that differently influence gender roles and expectations in the two nations.
Incorporating more balanced gender representations in textbooks is a matter of equity and a strategic necessity to inspire the next generation of women leaders in STEM fields (Van Craeynest 2015; Bachore 2022). The underrepresentation of women in educational materials sends a message to young girls that STEM is not for them, which further widens the gender gap in these critical fields. Textbooks are pivotal educational tools that shape students’ understanding and classroom dynamics (Davey 1988; Chiappetta and Fillman 2007; Andersson-Bakken et al. 2020; Orfan 2023). They socialize young learners, reinforcing societal values and gender norms (Lee and Collins 2010; UNESCO 2020). Gender representation in textbooks can perpetuate biases or promote equity, influencing students’ academic engagement and career aspirations (Potter and Rosser 1992; Good et al. 2010). Cultural perceptions significantly influence textbook content. China and the U.S. face common challenges in using educational resources to achieve gender equality. However, Chinese science textbooks may exhibit more stereotypes than American textbooks due to the influence of Confucian patriarchal values, which can be even more detrimental compared to the general patriarchal influences in the U.S. (UNFPA 2020). Therefore, examining gender portrayals in educational materials is crucial.
Textbooks also serve as sources of visual information, aiding comprehension through various visual data (Soesilo and Munthe 2020; Suryanda et al. 2020). However, the representation of gender roles in textbook images is poorly understood. Traditional content analysis methods are labor-intensive and limited in scope (Szasz et al. 2022). This study addresses these challenges using large language model (LLM) technology, a type of artificial intelligence that processes and understands human language, to analyze gender representation in STEM textbook images. In this study, we investigate gender representation in science textbooks from China and the U. S. using a large language model to analyze the implicit meanings behind images. The rationale for using an LLM includes its advanced capabilities in understanding and generating human-like text, allowing for nuanced visual content analysis. Traditional methods often fail to capture the complex, context-dependent meanings conveyed through images, especially in educational materials where subtle cultural and gender norms are embedded. Studies have demonstrated the effectiveness of LLMs in identifying and mitigating biases in vision-language models, providing a reliable framework for analyzing gender bias in images (Fraser and Kiritchenko 2024; Sathe et al. 2024).
Through a detailed comparative analysis of gender portrayals in middle school textbooks, the study identifies significant disparities in how genders are depicted. Findings reveal that Chinese textbooks show less gender balance than their U.S. counterparts and a significant underrepresentation of women in Chinese science textbooks. Limited research specific to Chinese gender stereotypes and inequalities in science-related textbooks aligns with these findings (Wang 2023; Yang and Zhou 2023). In contrast, the near parity in gender representation within U.S. science textbooks reflects efforts toward achieving equal representation in educational content, possibly linked to specific U.S. policies, e.g., Title IX. The role distributions of males and females in science textbooks from China and the U.S. show distinct gender diversity differences. Most roles are predominantly male, except for students, with a significant underrepresentation of females in key scientific roles. In the representational analysis, comparing the distribution of gender roles in science textbooks from China and the U.S., how cultural differences impact textbook content is evident. This suggests that gender roles are portrayed more equitably in U.S. textbooks, which serve as a conduit for cultural values and social norms.
The findings underscore the impact of cultural and educational norms on textbook content, revealing that while Chinese textbooks often align with traditional gender roles, U.S. textbooks tend to promote a more equitable view of gender roles in science. This study contributes to the academic discourse on gender bias in educational materials. It suggests ways to develop textbook content that fosters gender equality, potentially influencing future science curriculum designs and policies. For educators and policymakers, particularly in China and the U.S., there is an urgent directive to scrutinize and revise educational materials to ensure they portray gender equitably. [References available by request]

Authors

• Jialu Zhao, Stanford University

• Christine Min Wotipka, Stanford University