This paper explores the influence of passive haptic feedback on presence and task performance using two important interaction metaphors. We compared direct interaction with the user's hand with interaction using a stylus. Twenty-four participants performed a simple selection task consisting of pressing buttons while playing a memory game, with haptic feedback and interaction metaphor as the independent variables. We measured task performance by computing errors and time between button presses. We measured presence with questionnaires and through a new method based on users' involuntary movements. Our results suggest that passive haptic feedback improves both presence and task performance. However, small but significant differences related to the interaction metaphor were only apparent when haptic feedback was not provided.