Preference for gender stereotypicality in artificial intelligence

Spielmann, Julia

Permalink

https://hdl.handle.net/2142/115868 Copy

Description

Title

Preference for gender stereotypicality in artificial intelligence

Author(s)

Spielmann, Julia

Issue Date

2022-06-15

Director of Research (if dissertation) or Advisor (if thesis)

Stern, Chadly

Doctoral Committee Chair(s)

Stern, Chadly

Committee Member(s)

• Cohen, Dov
• Fairbairn, Catharine
• Todd, Nathtan
• Miller, Andrea

Department of Study

Psychology

Discipline

Psychology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• gender stereotyping
• artificial intelligence

Abstract

Do people prefer for voice-based artificial intelligence (AI) to align with gender stereotypes? Gender is a salient and ubiquitous organizing principle that provides a sense of structure and simplicity to society. Initial evidence suggests that non-human entities, such as weather patterns, numbers, robots, and even AI, are perceived within a human-gender context. However, no research to date has investigated whether people prefer gender stereotypicality in AI. AI provides a unique context to examine gender stereotyping that is removed from both social (e.g., gender discrimination laws that apply to humans) and non-social factors (e.g., biological influences on gender roles among humans). Across four studies using experimental designs, I examined whether people prefer voice-based AI that aligns with gender stereotypes. Across Studies 1 and 2, I found that people preferred gender stereotypicality (over counterstereotypicality and androgyny) in voice-based AI. In Studies 3 and 4, I found that people perceived AI as more credible when the gender of the AI voice was a stereotypical (versus counterstereotypical) match to the gender of the question, which in part explained people’s preference for stereotypicality. These studies contribute to understanding whether people prefer gender stereotypicality outside human gender relations, which holds implications for how AI might be used to both create and reinforce a gendered social reality.

Graduation Semester

2022-08

Type of Resource

Thesis

Copyright and License Information

Copyright 2022 Julia Spielmann

Owning Collections