ADULTS AND CHILDREN WITH HIGH IMAGERY SHOW MORE PRONOUNCED PERCEPTUALPRIMING EFFECT

36 children in Grade 5 and 59 university students, all native speakers of Japanese, studied three types of priming stimuli in a mixed list: words written in hiragana (Japanese syllabary used in writing), words written in Kanji (Chinese characters also used in writing), and pictur es. They were then given a task involving completion of hiragana-word fragments: the task involved studied and nonstudied items. For both ch ildren and university students, words in hiragana produced the largest priming effects, that is, the words that had appeared in hiragana in the preceding study phase were generated more often in the test phase of word completion than the other two types of priming stimuli. This c onfirms that the perceptual priming effect depends much on data-driven processing. For both age groups, words in Kanji produced nearly half the priming effects seen for hiragana-words. On the other hand, pictur es had no priming effect for children but they had a similar effect to Kanji-words for students. The discrepancy between Kanji-words and pic tures for children suggests that the former force the subject to read the words, which, possibly, activates the hiragana-words, while the la tter do not necessarily force labelling the pictures. Among three kind s of imagery tests, the Verbalizer-Visualizer Questionnaire predicted priming scores for children and the Questionnaire upon Mental Imagery did so for students, but the Test of Visual Imagery Control did not pr edict the scores for either age group. This shows that children report ing habitual use of imagery and adults reporting vivid imagery have mo re pronounced perceptual priming effects. We conclude that the imagery ability based on self-judgments reflects real characteristics of the perceptual representation system of Tulving and Schacter (1990).