PURPOSE: Measures of narrative quality hold promise for clinical language assessment. However, more information is needed on the validity of such measures within clinical populations. This study examined aspects of validity for two clinically-available narrative quality measures, the Narrative Scoring Scheme and a holistic scoring method, within a diverse group of children referred for speech-language assessment. METHOD: Assessment records from 72 school-age children were retrospectively analysed. Narrative quality scores from oral narrative language samples were examined for expected qualities including growth with age, similarity across groups defined by gender and linguistic background and relations with other measures of language ability including norm-referenced test scores and microstructural language sample measures. RESULT: External validity was supported by unbiased scores across gender and linguistic background. Concurrent, criterion-related validity was supported by strong correlations between narrative quality and some microstructural measures, as well as by the similarity in performance between the two tools. However, the expected relation with age was not present and there were few correlations between narrative quality scores and norm-referenced language tests, showing limitations in construct and criterion-referenced validity. CONCLUSION: Results provide an independent validation of these two tools and may guide clinicians looking to assess narrative quality.
PURPOSE: Measures of narrative quality hold promise for clinical language assessment. However, more information is needed on the validity of such measures within clinical populations. This study examined aspects of validity for two clinically-available narrative quality measures, the Narrative Scoring Scheme and a holistic scoring method, within a diverse group of children referred for speech-language assessment. METHOD: Assessment records from 72 school-age children were retrospectively analysed. Narrative quality scores from oral narrative language samples were examined for expected qualities including growth with age, similarity across groups defined by gender and linguistic background and relations with other measures of language ability including norm-referenced test scores and microstructural language sample measures. RESULT: External validity was supported by unbiased scores across gender and linguistic background. Concurrent, criterion-related validity was supported by strong correlations between narrative quality and some microstructural measures, as well as by the similarity in performance between the two tools. However, the expected relation with age was not present and there were few correlations between narrative quality scores and norm-referenced language tests, showing limitations in construct and criterion-referenced validity. CONCLUSION: Results provide an independent validation of these two tools and may guide clinicians looking to assess narrative quality.
Entities:
Keywords:
Narrative; assessment; school-age children