process where the observer sees the model punished, making the observer less likely to imitate the model’s behavior