A new paper published in Social Networks

Excited to share our latest open access study that was just published in Social Networks by Thijmen Jeroense, Jochem Tolsma, and myself!

📖 Find the paper here.

🔭What do we study?

• We ran an experiment with thousands of people in the Netherlands as to how they choose political discussion partners
• We consider dyadic similarity and extended networks using network scale-up methodology

🧐 What do we find?

• The overwhelming majority of people in the Netherlands (>70%) disclose political opinions or engage in substantive debate
• Men do so more often than women, as do theoretically educated rather than practically educated people
• Homphily on gender, ethnic background, or political opinion similarity surprisingly has little effects
• Whether you know someone well seems more important than whether you look alike (on gender, ethnic background, political opinion)

This was a very cool paper to write and combines complicated Bayesian NSUM models and vignette experiments with a very important sociological topic – i.e., discussing political may be a vehicle to de-polarize society. Here is the replication website to the paper.