Related Experiment Video
Updated: Jul 3, 2025

Post-Movie Subliminal Measurement PMSM, for Investigating Implicit Social Bias
Published on: February 29, 2020
Online images amplify gender bias
Douglas Guilbeault1, Solène Delecourt2, Tasker Hull3
1Haas School of Business, University of California, Berkeley, Berkeley, CA, USA. douglas.guilbeault@haas.berkeley.edu.
Online images significantly worsen gender bias compared to text. This visual shift amplifies stereotypes, particularly affecting women's representation and perception in various social categories.
Area of Science:
- Social Psychology
- Digital Media Studies
- Computer Science
Background:
- Increasing online image consumption over text.
- Visual content dominates social media and news.
- Images are processed faster and more memorably than text.
Purpose of the Study:
- To investigate how the rise of online images affects gender bias.
- To compare gender bias prevalence in images versus text.
- To assess the psychological impact of visual search on gender stereotypes.
Main Methods:
- Analysis of over one million images and billions of words from Google, Wikipedia, and IMDb across 3,495 social categories.
- Comparison of gender bias in images and text data.
- A nationally representative experiment on the effect of image-based vs. text-based occupational searches.
Main Results:
- Gender bias is consistently more prevalent in images than in text for both male- and female-typed categories.
- Online image underrepresentation of women is significantly worse than in text, public opinion, or census data.
- Searching for images of occupations amplifies gender bias in beliefs more than text searches.
Conclusions:
- The shift to visual communication online exacerbates gender bias.
- Addressing visual gender bias is crucial for an inclusive internet.
- Interventions are needed to mitigate the amplified stereotypes in visual online content.
Related Concept Videos
Stereotypes, Prejudice, and Discrimination
Socioemotional Experience and Gender Development
Bullying
Stereotype Threat and Self-fulfilling Prophecies
Confirmation Biases
Stereotype Content Model

