Gender bias in video game dialogue

Area of Science:

• Computational linguistics
• Media studies
• Gender studies

Background:

• Gender biases in dialogue are prevalent across media like film, TV, and books.
• Female characters often speak less, interact less with each other, and have narrower conversational topics.
• Video games, a major media form, lack comprehensive data on gender dialogue representation.

Purpose of the Study:

• To establish the first large-scale, consistently coded corpus of video game dialogue.
• To enable quantitative measurement and monitoring of gender representation in video game narratives.
• To identify specific gender biases in dialogue quantity, interaction, and content within video games.

Main Methods:

• Creation of the Video Game Dialogue Corpus, a novel dataset for analyzing game scripts.
• Systematic coding of dialogue to quantify representation across genders.
• Comparative analysis of dialogue volume, speaking partners, and thematic content for male and female characters.

Main Results:

• Female characters have approximately half the amount of dialogue compared to male characters.
• Dialogue disparities are attributed to both the underrepresentation of female characters and biases in their interactions and speech content.
• Analysis reveals specific patterns of bias in who female characters speak to and the topics they discuss.

Conclusions:

• The Video Game Dialogue Corpus provides crucial data for understanding gender representation in games.
• Significant gender biases exist in video game dialogue, impacting how female characters are portrayed.
• Recommendations are offered to game developers for creating more inclusive and equitable game narratives.