BACKGROUND: This study examined the narrative skills of Cantonese-speaking school-age children to fill a need for a normative language test for school-age children. PURPOSE: To provide a benchmark of the narrative skills of Cantonese-speaking children; to identify which of the microstructure components was the best predictor of age; and to determine the diagnostic accuracy of the test components. METHOD AND PROCEDURE: Data were collected from 1,120 Cantonese-speaking children between the ages of 4;10 (years;months) and 12;01, using a story-retell of a 24-frame picture series. Four narrative components (syntactic complexity, semantic score, referencing, and connective use) were measured. OUTCOMES AND RESULTS: Each measure reflected significant age-related differences in narrative ability. Regression analyses revealed that vocabulary and syntactic complexity were the best predictors of grade. All measures showed high sensitivity (86%-94%) but relatively low specificity (60%-90%) and modest likelihood ratio (LR) values: LR+ (2.15-9.42) and LR- (0.07-0.34). CONCLUSION AND IMPLICATIONS: Narrative assessment can be standardized to be a reliable and valid instrument to assist in the identification of children with language impairment. Syntactic complexity is not only a strong predictor of grade but was also particularly vulnerable in Cantonese-speaking children with specific language impairment. Further diagnostic research using narrative analysis is warranted.
BACKGROUND: This study examined the narrative skills of Cantonese-speaking school-age children to fill a need for a normative language test for school-age children. PURPOSE: To provide a benchmark of the narrative skills of Cantonese-speaking children; to identify which of the microstructure components was the best predictor of age; and to determine the diagnostic accuracy of the test components. METHOD AND PROCEDURE: Data were collected from 1,120 Cantonese-speaking children between the ages of 4;10 (years;months) and 12;01, using a story-retell of a 24-frame picture series. Four narrative components (syntactic complexity, semantic score, referencing, and connective use) were measured. OUTCOMES AND RESULTS: Each measure reflected significant age-related differences in narrative ability. Regression analyses revealed that vocabulary and syntactic complexity were the best predictors of grade. All measures showed high sensitivity (86%-94%) but relatively low specificity (60%-90%) and modest likelihood ratio (LR) values: LR+ (2.15-9.42) and LR- (0.07-0.34). CONCLUSION AND IMPLICATIONS: Narrative assessment can be standardized to be a reliable and valid instrument to assist in the identification of children with language impairment. Syntactic complexity is not only a strong predictor of grade but was also particularly vulnerable in Cantonese-speaking children with specific language impairment. Further diagnostic research using narrative analysis is warranted.