[Paper Review] Cobbler Stick With Your Reads: People's Perceptions of Gendered Robots Performing Gender Stereotypical Tasks

Sven Y. Neuteboom, Maartje M.A. de Graaf|arXiv (Cornell University)|Apr 13, 2021

Social Robot Interaction and HRI40 references4 citations

TL;DR

This study investigates how robot gender and task type influence human perceptions of trust, social perception, and humanness in an online experiment (n=89). It finds that task type—particularly analytical vs. social tasks—has a stronger effect on perceived competence and trust than robot gender, with female robots showing a dehumanization trend toward animals and gender-stereotype-violating robots toward emotionless objects.

ABSTRACT

Previous research found that robots should best be designed to fit their given task, whilst others identified gender effects in people's evaluations of robots. This study combines this knowledge to investigate stereotyping effects of robot genderedness and assigned tasks in an online experiment (n = 89) manipulating robot gender (male vs. female) and task type (analytical vs. social) in a between subject's design in terms of trust, social perception, and humanness. People deem robots more competent and have higher trust in their capacity when they perform analytical tasks compared to social tasks, independent of the robot's gender. Furthermore, we observed a trend in the data indicating that people seem to dehumanize female robots (regardless of task performed) to animals lacking higher-level mental processes, and additionally that people seem to dehumanize robots to emotionless objects only when gendered robots perform tasks contradicting the stereotypes of their gender.

Motivation & Objective

• To investigate the impact of robot gender and task type on human perceptions of trust, social perception, and humanness.
• To examine whether gender stereotypes influence evaluations of robots performing gender-stereotypical tasks.
• To explore dehumanization trends in human-robot interactions, particularly when robots violate gender-stereotypical role expectations.
• To assess whether perceived competence and trust in robots are more influenced by task type than by robot gender.
• To contribute to the understanding of implicit gender biases in human-robot interaction and their implications for robot design and societal norms.

Proposed method

• Conducted a between-subjects online experiment with 89 participants.
• Manipulated two independent variables: robot gender (male vs. female) and task type (analytical vs. social).
• Measured outcomes using validated scales for trust, social perception, and perceived humanness.
• Used statistical analysis to test for main effects and interaction effects between robot gender and task type.
• Employed stimuli with gendered robot appearances and gender-stereotypical task descriptions to elicit social categorization and stereotyping.
• Explored potential biases in stimuli perception, noting that robot gender and task ratings leaned toward the opposite end of the gender/task spectrum.

Experimental results

Research questions

• RQ1 How does robot gender influence people’s trust in robots performing analytical versus social tasks?
• RQ2 To what extent do gender stereotypes affect perceptions of robot competence and humanness?
• RQ3 Are there dehumanization effects when robots perform tasks that contradict their assigned gender stereotypes?
• RQ4 Does the perceived humanness of a robot vary based on the interaction between its gender and the task it performs?
• RQ5 How do people’s social perceptions of robots differ when the robot’s gender and task align or conflict with traditional gender role expectations?

Key findings

• People trust robots more when they perform analytical tasks compared to social tasks, regardless of the robot’s gender.
• Perceived robot competence is significantly higher for analytical tasks than for social tasks, with no significant effect from robot gender.
• A trend in the data indicates that female-identified robots are dehumanized to animals, suggesting a perceived lack of higher-level mental processes.
• When gendered robots perform tasks that contradict their stereotypical roles, people tend to dehumanize them to emotionless objects.
• The hypothesis that robot gender and task type jointly affect perceived humanness was not supported by the data, though a trend suggests potential interaction effects.
• The results suggest that task type exerts a stronger influence on human perceptions than robot gender, challenging assumptions about the primacy of gender in human-robot interaction.