This study tested the influence of the variables number of experimenters (one or two) and type of context (with game or with no-game) in children's performance in conservation task. Ninety six 5-7 years old children from a Rio de Janeiro's school were randomly distributed between 4 conditions and tested individually on a task involving conservation of discontinuous quantity. The conditions were: game context and one experimenter, game context and two experimenters, two experimenters and no-game context, and classical context (one experimenter and no-game). The results were analyzed by the Logistic Regression and offered clear evidence of the increasing frequency of correct responses in the groups with game context with 1 or 2 experimenters, and in the no-game context with 2 experimenters, but no interaction was observed. Possible problems with the procedures regularly employed in piagetian researches are discussed.
Conservation; Context; Child; Number of experimenters; Game