Is the evaluation of performance gender specific? Evidence from two large experimental studies

Area of Science:

• Social Psychology
• Gender Studies
• Human-Computer Interaction

Background:

• Societal assumptions suggest women receive less recognition than men for equivalent work.
• Previous research indicates presenter attractiveness can positively influence performance evaluations.

Purpose of the Study:

• To empirically test the assumption of gender bias in performance recognition.
• To investigate the role of presenter gender, attractiveness, and topic association in evaluation.
• To compare findings between artificial intelligence (AI)-generated avatars and real human presenters.

Main Methods:

• Two large-scale online survey experiments (N=3,157 and N=2,909) were conducted.
• Participants evaluated video presentations by either male or female characters (avatars or real humans).
• Presentation topics (male-associated, gender-neutral, female-associated) and presenter attractiveness were manipulated.

Main Results:

• Gender of the presenter did not significantly influence evaluation outcomes in either study.
• In Study I (AI avatars), attractive male presenters received more favorable evaluations than less attractive males; female avatars did not receive an "attractiveness bonus."
• In Study II (real humans), the attractiveness bonus observed for male avatars was not replicated for real male presenters.

Conclusions:

• The study found no evidence of gender bias in the evaluation of presentations.
• Presenter attractiveness influenced evaluations of male AI avatars but not female avatars or real human presenters.
• Findings suggest that AI-generated avatars may be perceived differently regarding attractiveness and its impact on performance evaluation compared to real humans.