The effect of partially valid cues on children's discrimination learning

Second grade children were administered a two-choice discrimination learning task in which irrelevant dimensions were correlated .50, .75, or 1.00 with the 100 percent rewarded cue. The results indicated that when a cue was partially correlated with reward .75, learning was retarded. When all cues were relevant and redundant (r = 1.00), learning was most rapid. These results supported an implication of Hypothesis Testing Theory that partially correlated or partially valid cues would not be eliminated rapidly, thus retarding acquisition of the correct response.