Cooperative social decision making among humans requires individuals to read the minds of their partners and estimate their moves. It also leads decision makers to focus on the value of mutual cooperation. In order to examine the distinctive characteristics of human-human decision making, we conducted an experiment that involved the manipulation of the game partner and explicitness of a partner's strategy independently. Participants played a prisoner's dilemma game with a computer partner or a human partner, whose decision making strategy was random or unknown. We observed that the participants increased their preference for mutual cooperation when their partners were human and the expectation for their partners' cooperation was raised when they were not informed about the partners' decision making strategies. Thus, it was found that the social context and expectation about the partner's decision making strategy resulted in different cognitive processes although both of them elicit cooperation.