Collection terms and preschoolers' use of the cardinality rule

Three experiments examined the effect of class (e.g., pigs) versus collection (e.g., pig family) terms in preschool children's giving of their last counting word in response to an inquiry as to how many objects there were in an array (the operational definition of the cardinality rule). In Experiment 1 children gave the last counting word to the how-many question as often when the array was described with class terms as with collection terms, and these terms did not interact with set size or homogeneity of appearance of the array objects to affect cardinality-rule performance. Experiments 2 and 3 also failed to show any effect of class vs collection terms on children's giving of last counting word responses to the how-many question. Counting accuracy in all three experiments also was unaffected by class-collection terms. Thus these experiments failed to replicate the collection term effect on the cardinality principle. Three lines of argument are advanced to suggest that this failure to replicate is limited to the cardinality-rule task and does not extend to set comparison tasks (e.g., conservation of number, class inclusion).