Purpose Many studies have found a correlation between overall usage rates of nonmainstream forms and reading scores, but less is known about which dialect differences are most predictive. Here, we consider different methods of characterizing African American English use from existing assessments and examine which methods best predict literacy achievement. Method Kindergarten and first-grade students who speak African American English received two assessments of dialect use and two assessments of decoding at the beginning and end of the school year. Item-level analyses of the dialect-use assessments were used to compute measures of dialect usage: (a) an overall feature rate measure based on the Diagnostic Evaluation of Language Variation-Screening Test, (b) a subscore analysis of the Diagnostic Evaluation of Language Variation-Screening Test based on items that pattern together, (c) an alternative assessment where children repeat and translate sentences, and (d) "repertoire" measures based on a categorical distinction of whether a child used a particular feature of mainstream American English. Results Models using feature rate measures provided better data-model fit than those with repertoire measures, and baseline performance on a sentence repetition task was a positive predictor of reading score at the end of the school year. For phonological subscores, change from the beginning to end of the school year predicted reading at the end of the school year, whereas baseline scores were most predictive for grammatical subscores. Conclusions The addition of a sentence imitation task is useful for understanding a child's dialect and anticipating potential areas for support in early literacy. We observed some support for the idea that morphological dialect differences (i.e., irregular verb morphology) have a particularly close tie to later literacy, but future work will be necessary to confirm this finding. Supplemental Material https://doi.org/10.23641/asha.13425968.

中文翻译：

---

非裔美国人英语和早期识字：量化非主流方言使用的方法比较。

目的 许多研究发现非主流形式的总体使用率与阅读分数之间存在相关性，但对于哪些方言差异最具有预测性却知之甚少。在这里，我们考虑从现有评估中表征非裔美国人英语使用的不同方法，并检查哪些方法最能预测识字成绩。Method 讲非裔美国英语的幼儿园和一年级学生在学年开始和结束时接受了两次方言使用评估和两次解码评估。方言使用评估的项目级分析用于计算方言使用的度量：（a）基于语言变异筛选测试的诊断评估的整体特征率度量，(b) 基于共同模式的项目的语言变异筛查测试的诊断评估的子分数分析，(c) 儿童重复和翻译句子的替代评估，以及 (d) 基于分类区分的“曲目”措施一个孩子是否使用了美国主流英语的特定特征。结果 使用特征率测量的模型比使用曲目测量的模型提供了更好的数据模型拟合，并且句子重复任务的基线表现是学年结束时阅读分数的积极预测因子。对于语音子分数，从学年开始到结束的变化预测了学年结束时的阅读，而基线分数最能预测语法子分数。结论 添加句子模仿任务有助于理解儿童的方言和预测早期识字支持的潜在领域。我们观察到一些支持的观点，即形态方言差异（即不规则动词形态）与后来的识字有特别密切的联系，但未来的工作将有必要证实这一发现。补充材料 https://doi.org/10.23641/asha.13425968。