Premium
The study of play and problem solving in preschool children: Have experimenter effects been responsible for previous results?
Author(s) -
Simon Tony,
Smith Peter K.
Publication year - 1983
Publication title -
british journal of developmental psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.062
H-Index - 75
eISSN - 2044-835X
pISSN - 0261-510X
DOI - 10.1111/j.2044-835x.1983.tb00901.x
Subject(s) - psychology , task (project management) , cognitive psychology , replication (statistics) , developmental psychology , social psychology , statistics , mathematics , management , economics
Several studies have suggested that, so far as lure‐retrieval tasks are concerned, previous ‘play’ and ‘training’ experiences are roughly comparable in facilitating subsequent problem‐solving performance. In certain circumstances, ‘play’ experience has proved superior. However, Cheyne (1982) has criticized all these studies for poor or absent controls for possible experimenter bias. The present study carried out a close replication of the Smith & Dutton (1979) experiment, but incorporated manipulations of and controls for experimenter bias. Even when the experimenter giving the problem task was unaware of the previous experience of the child, and when scoring was done blindly, it was found that the effects of the ‘play’ experience were equivalent to those of a ‘training’ experience directly related to the problem to be given. This replicated previous results but excludes experimenter bias as a possible explanation. However, the second finding of the Smith & Dutton (1979) study, that ‘play’ experience was markedly superior to the ‘raining’ experience for a problem task more innovative with respect to the training, was not replicated. It is hypothesized that a combination of procedural differences and possible experimenter bias as suggested by Cheyne (1982) was responsible for this latter result.