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

PROCEDURES

Participants in both grades were administered the mora deletion, RAN, nonword repetition, visual search, and oral reading fluency measures in individual sessions lasting 20 to 30 minutes. To avoid giving similar tasks consecutively, the following sequence was used: (1) number RAN, (2) mora deletion, (3) object RAN, (4) oral reading fluency, (5) hiragana RAN, (6) visual search, (7) kanji RAN (first graders only), and (8) nonword repetition. The silent reading comprehension test was group administered to the first graders in their classroom. Students were given 20 minutes to complete this test.

RESULTS

Table I provides mean raw scores for the processing variables and reading measures administered to kindergartners and/or first graders. Independent t-test comparisons across grades indicate significantly higher scores for first graders on all but one of the processing variables: mora deletion, object RAN, number RAN, hiragana RAN, visual search errors, and visual search speed. Effect sizes indicate that all significant differences are moderate to strong (.74-1.56). Thus, with the exception of nonword repetition, which did not differ significantly between groups, the processing measures appear to be developmentally sensitive. Cross-grade comparisons of performance on the reading measures are not appropriate because the kindergartners and first graders read different passages.

Kindergartners. The next set of analyses examined correlations among the processing variables and reading measures for the kindergarten group. Correlations are presented in the lower diagonal in table II. A highly significant correlation was found between number RAN and hiragana RAN (r = .86). Correlations among the other processing variables were not significant, with the exception of a moderate correlation between mora deletion and nonword repetition (r = .44). In terms of reading measures, number RAN and hiragana RAN correlated significantly with both oral reading accuracy and speed. As one might expect, reading accuracy and speed were highly inter-correlated (r = .72).

A stepwise multiple regression analysis (probability-of-F-to-enter .10) was used to determine the best set of predictors of reading speed and accuracy among the processing variables (mora deletion, nonword repetition, object RAN, number RAN, hiragana RAN, visual search speed, and visual search errors). The left side of table III shows the results with reading accuracy as the dependent measure. The overall R-square was .67 (p = .002) and the only significant predictor was hiragana RAN. The right side of table III shows the results with reading speed as the dependent measure. The overall R-square was .83 (p

First Graders. Correlations among processing variables and reading measures for the first graders are shown in the upper diagonal of table I. All RAN tasks were significantly intercorrelated; the correlation between number RAN and kanji RAN was the highest (r = .70). Both mora deletion and visual search speed correlated significantly with object RAN and kanji RAN. Visual search errors and nonword repetition did not correlate with other processing variables. The variable most highly correlated with oral reading accuracy was mora deletion (r = .53), followed by number RAN (r = .38) and kanji RAN (r = .34). Six of the eight processing variables correlated significantly with oral reading speed: the four RAN tasks, mora deletion, and visual search speed. All of the processing variables except visual search errors correlated significantly with reading comprehension. All three reading measures were significantly inter-correlated.