ABSTRACT
The study conducted with 51 nurse students produced three results. First, the students who studied in pairs built more abstract representations of clinical cases than those working individually. Second, the fact of building a more abstract representation did not lead to higher scores in a transfer task. Third, the students who were trained with examples taken from varied contexts got higher scores in a transfer task than those working with examples from the same context. The partial contradiction between these results questions the relationship between abstraction and transfer. It may also indicate that the abstract representations built as a side-effect of grounding mechanisms were not internalized into personal representations.
