Using Discourse Analysis Tools to Describe Academic Language Produced by K-3 Students

• In Event: Academic Language of PreK to 5th Grade Students: Innovative Measurement and Broad Conceptualization

Abstract

Academic language is specialized language, in form and function, that consists of discourse-, sentence-, and word-level patterns. It is often referred to as the language of schooling or the language needed to acquire and express knowledge. Although there have been several empirical investigations of academic language understood or produced by upper elementary and intermediate students, there is relatively no research that has documented the academic language of younger students. Without an understanding of the developmental progression of specific features or clarity about how academic language differs across grades, modalities (e.g., oral vs. written) and registers (e.g., narrative vs. expository), the creation of assessment and instructional tools is hindered. Therefore, the purpose of this research was to catalog the academic language produced by primary grade students by examining multiple dimensions of academic language in relation to their elicitation context.

In a stratified cohort corpus linguistic study, 8826 academic language samples were collected from 1,040 diverse kindergarten (n=282), first grade (n=257), second grade (n=279), and third grade (n=222) students. The student sample consisted of 28.8% White, 25.4% African American, 27.1% Hispanic, 13.7% Biracial, 2.5% Asian, <1% Native American, and 2.5% other students. 13.5% of students had an IEP at the time of enrollment in the study and 20% spoke a language other than English at home.

Examiners engaged participants in three individual sessions in a quiet location within elementary school buildings and all sessions were audio recorded. With a few exceptions, each participant contributed eight oral language samples and one written sample. A set of standardized elicitation materials and procedures were used to collect oral and written samples across four specific contexts: narrative retell, narrative generation, expository retell, and expository generation. A highly trained research team scored all the samples using a set of discourse analysis tools (i.e., Narrative Language Measures [NLM] Flowchart and Expository Language Measures [ELM] Flowchart; Spencer & Petersen, 2020) to describe participants’ inclusion of discourse-level structures, complex sentences, and word usage.

Data will be presented as means, standard deviations, and ranges for each grade level. In addition to graphical displays of these data, statistical analyses will be summarized in tables. Results indicate that discourse completeness, sentence complexity, and variety of word usage for both narrative and expository language, and for both oral and written samples, increase as a function of grade. Overall, narrative samples contain more causal and temporal subordinate clauses than expository samples, whereas expository samples contain more relative clauses than narrative samples. Although the length of samples differs across oral and written modalities, the complexity of the language students produce is generally the same across their oral and written samples.

Authors

• Trina D Spencer, University of South Florida
  Presenting Author

• Courtney Claar