Age differences in hypothesis testing and frequency processing in concept learning

Abstract

Two types of processes, controlled hypothesis testing and automatic frequency processing, have been posited to explain concept learning. The present study employed a concept-identification task to compare fifth-graders, college students, and elderly adults in their ability to test hypotheses and compile feature frequency information. College students proved superior in both their efficiency of selecting the correct hypothesis and their accuracy of recalling sampled hypotheses, and fifth-graders and elderly adults performed about the same on these measures of hypothesis-testing ability. Ability to estimate feature frequencies followed the same developmental pattern. The results support the view that hypothesis testing is an effortful, controlled process, but they contradict the assumption that frequency processing is automatic.