Ecologists use null models every day. But we rarely use them when measuring gender bias in academia.