Development of the middle concept.

Children's use of the middle concept was assessed in two developmental studies. In Experiment 1, children in kindergarten through Grade 5 were given middle-size pretraining followed by the presentation of sets of three stimuli representing eight novel dimensions. There was a marked improvement in the mastery of the middle concept across elementary school grades. Asking the child to point to the middle "dimension label" (e.g., middle oval) facilitated performance when compared to asking the child to point to the middle thing. In Experiment 2, second- through fifth-grade children were pretrained to choose the middle item associated with a single set of three stimuli, two sets of three stimuli representing the same dimension, or two sets of three stimuli with each set representing a different dimension. Transfer to six novel test dimensions was evaluated. Pretraining with two rather than one stimulus set transferred to the novel test dimensions. The results were related to induction and transposition theories and to issues concerning implicit and explicit learning.