Abstract
Growing evidence suggests that infants and young children distinguish between prosocial and antisocial others. The main paradigm employed to test this ability is the hill paradigm. The hill paradigm involves three characters: a protagonist trying to climb a hill, an agent that facilitates them to achieve their goal by pushing them up the hill (i.e., the Helper), and an agent that prevents them from achieving their goal by pushing them down the hill (i.e., the Hinderer). After watching these scenarios infants and young children not only display a preference for the Helper, but also hold expectations about how the protagonist should interact with the Helper and Hinderer. The first aim of the current study was to replicate these basic findings with both children and young adults. Specifically, we hypothesised that when the Helper and Hinderer were presented side-by-side, participants would show a looking preference for the Helper (Hypothesis 1a) and expect the protagonist to approach the character that helped, rather than hindered, its ascent of the hill (Hypothesis 1b). In addition, when participants’ own helping behaviour was assessed, we expected the level of helping they provided to correlate with their preference and expectation looking data (Hypothesis 2). Finally, we investigated how participants’ preferences and expectations following the hill paradigm, and their own helping behaviour, correlated with psychopathic traits. Specifically, we assessed social cue focus (i.e., face looking preference, especially eyes), and Callous-Unemotional (CU) Traits, hypothesising that those with a higher incidence of CU traits would have improved memory for the Helper (Hypothesis 3a) and that there would be a relationship between social cue focus and participants’ preferences and expectations following the hill paradigm (Hypothesis 3b). Our results revealed that none of the hypotheses were supported; participants looked equally at the Helper and Hinderer (Hypothesis 1a) and did not expect the protagonist to approach the Helper (Hypothesis 1b). Further, participants’ own helping behaviour did not correlate with the measures derived from the hill paradigm (Hypothesis 2), nor did their social cue focus (Hypothesis 3a) or their incidence of CU traits (Hypothesis 3b). These results call into question the basic findings derived from the hill paradigm and the assumption it is a marker of prosocial behaviour or preference.