Recent experiments(Magara, 1989) demonstrated that a high typicality instance(the pig) showed more positive effects than a low typicality instance (the silk worm) on learning from reading materials describing some properties of the livestock. And such results were discussed from the viewpoint of “typicality effect”. The purpose of the present paper was to find an explanation to the methodological problem in original researches, and to confirm again “typicality effect”. The obtained results failed to come to a proper replication. The original findings might be accounted for by the fact that most subjects are inclined to place high typicality instances rather than low ones in the category concerned in their pre-existing knowledge.