Effects of Mora Deletion, Nonword Repetition, Rapid Naming, and Visual Search Performance on Beginning Reading in Japanese

Annals of Dyslexia, Jun 2005 by Kobayashi, Maya Shiho, Haynes, Charles W, Macaruso, Paul, Hook, Pamela E, Kato, Junko

The main predictors of first graders' reading speed were RAN tasks: kanji and hiragana. Findings for hiragana RAN are consistent with the kindergarten results. For first graders, kanji RAN (not used with kindergartners) proved to be the best predictor of reading speed. As in the kindergarten results, number RAN was highly correlated with hiragana RAN (.69) and kanji RAN (.70), and when the latter predictors were removed from the regression model, number RAN became the strongest predictor of reading speed (accounting for 36% of the variance). Note that in contrast to kindergartners, object RAN in first graders is also significantly correlated with reading speed. This is not surprising given some of the similarities between the stimuli used in object RAN and kanji characters. Object-naming tasks employ pictures, and kanji characters are ideographic and sometimes pictorial in nature. First graders' speed of object naming might, therefore, capture some of the same processes involved in timed reading of text that includes kanji. Support for this argument can be found in the high correlation between object RAN and kanji RAN performance in first graders (.64), compared to the lower correlation between object RAN and hiragana RAN (.30).

With respect to reading comprehension in first graders, significant predictors included three of the four types of processing variables: kanji RAN, mora deletion, and nonword repetition. This finding is consistent with studies examining cognitive-linguistic factors and reading in English-speaking children (e.g., Wolf, 1991), and underscores the complex and multifaceted requirements of reading comprehension. It is noteworthy that after kanji RAN, mora deletion was the second strongest predictor of reading comprehension. As noted above, mora deletion was an important contributor to accuracy of oral reading in first graders and, therefore, its effect on comprehension could be related to its role in word recognition. At the same time, the significance of nonword repetition in predicting reading comprehension might reflect the key roles of phonological span and phonological working memory in processing sentences and more extended text (see Shankweiler & Grain, 1986; Swanson, 1999).

Overall, performance on the visual search task contributed only slightly to predicting reading. No significant relationships were found in the kindergarten data. Thus, at the earliest stages of reading in Japanese, RAN and mora deletion tasks (both requiring activation of phonological information) are more significant predictors of reading success than visual processing. However, significant correlations among visual search speed, object RAN, kanji RAN, reading speed, and reading comprehension were found in the first grade data. In fact, when kanji RAN is removed from the regression analysis, visual search speed becomes a significant predictor of reading comprehension, accounting for an additional 10% of the variance after mora deletion and nonword repetition are included in the model. These findings are consistent with previous reports that visual processing contributes to prediction of reading skills both in English and Chinese orthographies (Adams, 1990; Badian, 1994; Bowers & Wolf, 1993; Ho et al, 2004). The timed elements of the visual search and RAN tasks might reflect automaticity and fluency in recognizing orthographic symbols while reading text. Given Badian's findings regarding the relationship between visual processing and reading fluency in English, it is likely that the effects of visual search speed on reading performance in Japanese is greater in higher grades when text is orthographically more complex (see Badian, 1995).