ひらがな文字の形態的類似性評定調査

抄録

The process of recognizing printed words has been studied for many years. Recent visual word recognition research suggests that the identification of a word is affected by its similarity to other words. Coltheart,Davelaar,Jonasson,and Besner (1977) defined an orthographic neighbor as any word that can be generated by replacing one letter of a word. Many studies have shown that neighborhood size affects visual word recognition. The figurative similarity of neighbors can be defined in terms of the figurative similarities of component letters. The purpose of this study is to develop figurative similarity norms for Japanese hiragana in order to consider a more stringent definition of neighbors and to investigate neighborhood effects for Japanese hiragana words. In this article,the figurative similarity of hiragana refers to subjective ratings made to pairings of hiragana. In the study,a set of 71 hiragana was selected,and these were paired to form 2,485 (＝71 x 70/2) hiragana pairings. The hiragana pairings were divided into 25 stimuli lists. The 25 lists were constructed by randomly arranging the stimuli. A second set of 25 lists was also constructed by reversing the order of items on the first set of lists. The resultant 50 stimuli lists were presented to 1,750 university students for subjective evaluations concerning the figurative similarity of each hiragana pairing on a 5-point scale,ranging from 1 (not similar) to 5 (very similar). The results are reported in Table 2. The table may be employed as normative figurative similarity data when conducting experiments involving Japanese hiragana words.