Statistics for Judge-target trait similarity and accuracy: Does it 'take one to know one'?