We replicate in the lab an artefactual field experiment originally run with a representative sample of the population. Our results show that, despite the many differences between university students and representative subjects from the whole population, the two samples closely follow a common behavioral pattern in a set of binary dictator games. The only exception seems to be represented by a significant difference in those situations where self-interest plays a prominent role. This gap is mainly related to the academic background of the participants: our sample of undergraduate economics students, in fact, differs in its degree of self-interested choices both from the representative group of the population and from its sub-sample of students from heterogeneous disciplines.