Lab versus Online Experiments: Gender Differences

Comparing the behaviour of subjects randomly allocated to perform tasks and tests online or in the lab, I find that the setting influences males’ and females’ behaviour differently. Males are more selfish in the online setting than in the lab, while females are more risk-averse when they answer online compared to the lab. Interestingly, performance in math is influenced by the setting for females: online they perform better. The gender gap in numeracy closes online, while it is present in the lab, where males solve more math problems. This work provides preliminary evidence on how gender differences and gaps can be altered by the setting. As online settings have become more and more common practice in running experiments and have started to be used also in education, it is important to understand the consequences that this shift implies also relative to gender differences and gaps.