Abstract
In order to examine the relative effects of using photographs and actual objects in training for generalization, a group of severely-moderately retarded adults was taught a simple industrial type task(n=24). Subjects were randomly assigned to one of four equal groups: A (Objects only); B (Photographs only); C (Objects and Photographs); D (Conventional teaching through modelling). It was found that, with respect to the initial teaming, generalization and maintenance stages, the use of photographs as stimulus materials produced inferior results to those obtained from using actual objects or conventional teaching.