ABSTRACT
The evidence that the context in which children are given cognitive tasks has a huge effect on their performance in those tasks is virtually irrefutable. There are now many completely convincing demonstrations that children perform much better in one condition, or context, than in another, despite the fact that both contexts make exacdy the same logical and cognitive demands. Usually the contexts which cause children the most difficulty are the traditional conditions in which some very well-known cognitive tasks, such as the conservation problem and the class inclusion problem, are administered. In other circumstances the same children often manage the same problem with much greater success (Rose and Blank, 1974; McGarrigle and Donaldson, 1974–5; Light, Buckingham and Robbins, 1979; Roazzi, 1986). Such contextual effects have also been found in educational tasks. For example, children who work in street markets in Brazil make arithmetical calculations more effectively at their market stalls than in their classrooms (Carraher, Carraher and Schliemann, 1985).
