Optional shift behavior as a function of dimensional preference and relative cue similarity

Dimensional preference (either color or form) was assessed in first- and third-grade children prior to their being presented with a reversal-nonreversal optional shift task. During initial learning of the reversal-nonreversal task, one-half of the children were trained with their preferred dimension relevant, while the other half were trained with their preferred dimension irrelevant. For half of each of these groups, the relevant cues were similar; for the other half, the irrelevant cues were similar. Initial learning was faster when the preferred dimension was relevant and when the irrelevant cues were similar. More reversal shifts were made when the irrelevant cues were similar.