Stimulus overselectivity by mentally retarded adolescents: effects of pretraining on cue identification

Severely and profoundly mentally retarded adolescents were given two-cue redundant relevant cue discrimination problems under each of three conditions. In one condition both cues were ones that subjects had been previously trained to associate labels with. In the second condition one pretrained and one untrained cue were used, and in the third, two untrained cues were employed. Results from test trials indicated that the amount of stimulus overselectivity was lowest in the third condition. There was a much greater amount of overselective responding in both the first and second conditions. The results offered little support for the hypothesis that teaching severely retarded learners to label stimuli would reduce the likelihood of stimulus overselectivity. The relationship of the results to previous findings concerning the effects of stimulus and training conditions on overselectivity was also discussed.