Behaviour Modification with Children: The Generalisation Trap

Abstract

In order to encourage research in generalisation, Drabman, Hammer, and Rosenbaum (1979) introduced a framework to categorise the various generalised effects of child behaviour therapy. Sixteen different potential classes of generalisation were identified. To ascertain the current status of generalisation in the child behaviour therapy literature and further encourage generalisation research, we reviewed articles that appeared in 28 journals over the past 12 years. Of 15,141 studies, only 424 involved children and presented data on generalisation. Results indicated that: (a) a small percentage of studies both involved children and presented data on generalisation (2.77%), (b) of these studies, most failed to meet our methodological criteria for demonstrating generalisation, (c) the generalisation map categories of time, maintenance, setting, and setting-time were the most frequently encountered, (d) there was a significant increase in reported instances of maintenance generalisation effects over the past 10 years when compared with data from our earlier paper, and (e) generalisation data were found concerning 15 of the 16 map classes. Implications of these data, methods of conceptualising generalisation, the generalisation trap, and antecedent strategies for promoting generalisation are discussed.