Themes, taxons, and trial types in children's matching to sample: methodological considerations

Five experiments were conducted in which preschool children (29-73 months) matched pictures of common objects. Each trial consisted of a sample and two comparison pictures (all three together defined as a trial type) in which the sample was related to one comparison taxonomically and to the other thematically. The intent of each experiment was to replicate results in the literature in which young children more readily relate objects thematically than taxonomically. Instead, in all five experiments the children more often selected the taxonomic comparisons. These results occurred independently of age, gender, instructions, order of the trial types, nonspecific performance feedback, the presence of a third comparison unrelated to the sample, and the level of the taxonomy (superordinate vs basic). Instructions to select a comparison that was the same as the sample increased responses to the taxonomic comparisons. Comparison selection in this task relates both to the individual trial type and the individual subject.