Fig. 2.

Trial structure for the moral judgment task. (A) Each trial consisted of three screens that began by indicating a default number of shocks and money an agent would receive if that agent did nothing (above). This was followed by an alternative number of shocks and money that the agent would receive if they decided to switch (below). Finally, the agent’s choice was revealed and participants judged each choice on a scale ranging from ‘blameworthy’ to ‘praiseworthy’. (B) Heat maps depicting the bad and good agents’ probability of choosing the more harmful option as a function of the money gained and shocks delivered. As the option becomes more profitable (i.e., as more money is offered at the cost of each shock), the probability of choosing the harmful option increases, but the bad agents require less profit to choose the harmful option than do the good agents.