A replication of a test of the contact hypothesis in a different panel data setting supports the original study’s finding that outgroup contact does not affect outgroup bias.

Menu