Relational skills—taught during the critical period of adolescence —can set vulnerable young women’s lives on better trajectories 10 years later, from Nava Ashraf, Natalie Bau, Corinne Low, and Xiaoyue Shan www.nber.org/papers/w34339

🏁 Gender-diverse peer groups can be powerful and beneficial.

Educators and institutions should consider forming mixed-gender rather than homogeneous groups, e.g., project teams, lab groups, shared offices, production teams, etc.

Read full paper here: dropbox.com/scl/fi/a2h15...

4⃣ Gender diversity significantly improves students' subjective well-being:
– Satisfaction with the assigned group ↑
– Self-confidence about future achievement ↑
– Mental health ↑

3⃣ Gender diversity increases men's general competitiveness, which associates with significantly higher performance.
– Groups are collaborative: men are not competing against group members
– Exposure to the opposite gender still makes them more competitive

2⃣ Gender diversity leads to more collaborative & inclusive social dynamics:
– Boosts group/pairwise activities, both within- & cross-gender (see fig)
– Especially interactions with lowest-performing member
→ Knowledge sharing, peer mentoring, or role model effects

So why does gender diversity improve performance?

1⃣ The effect is not driven by & distinct from other diversities:
– Robust to controlling for a large set of group characteristics
– Diversities of nationality, cognitive/non-cognitive ability have different effects (see fig)

Spillovers:

Impact of gender diversity extends beyond the course where study groups are assigned.
It also …
▶️ improves grades in other first-semester courses
▶️ improves grades in next-semester courses, especially compulsory ones
▶️ lowers dropout rates one year later

Which member benefits most?

Group-level analyses show that gender diversity
▶️ significantly raises group minimum grade
▶️ has no impact on maximum grade
▶️ substantially reduces grade variance

Implication:
→ No one gets hurt
→ Academically struggling member benefits most

Main finding:

Students assigned to a more gender-diverse group perform significantly better.

▶️ Moving from a homogeneous to a gender-balanced group raises grades by 0.15 SD (p-val=.003)
▶️ Similar effect for ♂️ & ♀️
▶️ Robust across baseline ability, personality, SES and 7 cohorts

Set-up:

3,060 students in a first-semester course at a European university are randomly assigned to 765 study groups
→ Exogenous & large variation in gender composition

Groups are designed for long-term collaboration in common courses
→ Strong connections & lasting effects

Does gender diversity in peer groups benefit performance?

I ran a 7-year field experiment with 3,000+ students to find out.

Key findings: As gender diversity increases
– Course grades in short & longer run ⬆️
– Dropout rates ⬇️
– Performance inequality in groups ⬇️

Details below🧵

🚨 New working paper! 🚨
We ran a field experiment with >1,300 university students randomly assigned to online vs in-person lectures. What happens? Online instruction hurts performance, but only for women♀️📉

with Xiaoyue Shan & Uschi Backes-Gellner
www.ifo.de/DocDL/cesifo...

Come work with us at the National University of Singapore. Every year, we have a fantastic group of pre-docs working with us! #econ_ra

fass.nus.edu.sg/ecs/pre-doc-...

@mdrouvelis.bsky.social and I co-edit a special section on experimental replications at European Economic Review. We are eagerly waiting for your replications of lab and online experiments. Please get in touch if you have any questions. #econsky

This paper sparked lively discussion in our seminar. Lots of excitement mixed with valid concerns.

arxiv.org/abs/2411.10109