| Abstract: | Forty retarded children, a Low Mental Age (MA) Group (mean MA 3-3 yrs.) and a High MA Group (mean MA 5–7 yrs.) were trained on 120 different two-choice visual discrimination problems. For all subjects, half of these problems were standard four-trial problems (Standard Condition). The remaining 60 problems were variants of the Moss-Harlow Design, the first trial differing from the initial two-choice trial of the Standard Condition in having only a single stimulus present, either the positive stimulus (Positive Condition) or the negative stimulus (Negative Condition), in both cases followed by three trials with both stimuli present. Each subject received either Positive Condition Problems or Negative Condition Problems, but not both. An analysis of Trial 2 performance revealed that although no interproblem changes in the Standard Condition occurred, there were large initial performance differences and large interproblem changes in the Positive and Negative Conditions which varied with MA. Initial performance differences were interpreted as a differential preference for novel and familiar stimuli in the two MA Groups. Changes in interproblem performance were consistent with the hypothesis that these MA dependent novelty-familiarity preferences could be influenced by novelty-familiarity redundancy training. These interpretations were supported by predictions from theory. |
