Language Variation in the Writing of African American Students: Factors Predicting Reading Achievement

Purpose This study aims to examine the predictive relation between measures obtained from African American students' written narrative language samples and reading achievement, as measured by standardized academic assessments. Method Written language samples were elicited from 207 African American students in Grades 1-8. The samples were examined for morphosyntactic variations from standardized written Generalized American English (GAE). These variations were categorized as either (a) specific to African American English (AAE) or (b) neutral across AAE and standardized written GAE (i.e., considered ungrammatical both in AAE and in standardized written GAE). Structural equation modeling was employed to then examine the predictive relation between the density of AAE-specific forms in students' writing and their performance on standardized assessments of literacy and reading vocabulary. This relation was examined while accounting for the density of dialect-neutral morphosyntactic forms, reported family income, age, and written sample length. Results The written samples were highly variable in terms of morphosyntax. Younger students and those from lower income homes tended to use AAE-specific forms at higher rates. However, the density of AAE-specific forms did not significantly predict standardized literacy scores or reading vocabulary after accounting for dialect-neutral variations, income, and sample length. Conclusions These results support the ongoing need to better understand the language, literacy, and overall academic development of students from all backgrounds. It may be essential to focus on dialect-neutral language forms (i.e., morphosyntactic forms that are consistent across both AAE and standardized written GAE) in written samples to maximize assessment validity across students who speak varying dialects of English. Supplemental Material https://doi.org/10.23641/asha.16879558.

Marking of Tense and Agreement in Language Samples by Children With and Without Specific Language Impairment in African American English and Southern White English: Evaluation of Scoring Approaches and Cut Scores Across Structures

Purpose As follow-up to a previous study of probes, we evaluated the marking of tense and agreement (T/A) in language samples by children with specific language impairment (SLI) and typically developing controls in African American English (AAE) and Southern White English (SWE) while also examining the clinical utility of different scoring approaches and cut scores across structures. Method The samples came from 70 AAE- and 36 SWE-speaking kindergartners, evenly divided between the SLI and typically developing groups. The structures were past tense, verbal -s, auxiliary BE present, and auxiliary BE past. The scoring approaches were unmodified, modified, and strategic; these approaches varied in the scoring of forms classified as nonmainstream and other. The cut scores were dialect-universal and dialect-specific. Results Although low numbers of some forms limited the analyses, the results generally supported those previously found for the probes. The children produced a large and diverse inventory of mainstream and nonmainstream T/A forms within the samples; strategic scoring led to the greatest differences between the clinical groups while reducing effects of the children's dialects; and dialect-specific cut scores resulted in better clinical classification accuracies, with measures of past tense leading to the highest levels of classification accuracy. Conclusions For children with SLI, the findings contribute to studies that call for a paradigm shift in how children's T/A deficits are assessed and treated across dialects. A comparison of findings from the samples and probes indicates that probes may be the better task for identifying T/A deficits in children with SLI in AAE and SWE. Supplemental Material https://doi.org/10.23641/asha.13564709.