Project Activities
The researchers recorded students telling stories or giving information in response to pictures, including generation and retelling tasks. They collected samples from students who were below average and above average in terms of their oral language abilities. The researchers then examined the recordings for vocabulary and grammatical features and documented differences between groups of students, elicitation contexts, and across grades.
Structured Abstract
Setting
This study took place in Florida.
Sample
The sample included 1,074 kindergarten through third grade students.
Factors
Researchers investigated the distinct vocabulary and grammatical features that students with above- and below-average oral language abilities use in various academic conditions. These features of academic language can be improved upon through instruction which is associated with later achievement in reading and writing. This study's findings provided critical information for developing interventions to improve academic language for early elementary school students. For example, researchers catalogued oral academic language related to narratives and exposition. This involved better understanding under what contexts oral language is most complex and the differences across groups of students and grades. The results led to the specification of academic language features that can be expected at each grade level and which indices can help identify students with language disabilities.
Research design and methods
This research took place in four phases. In phase 1, the researchers developed a set of photo stimuli and scripted procedures for eliciting both narrative and expository academic language samples from students. In phase 2, they used the photos and procedures to gather narrative and expository academic language samples from 1,074 diverse kindergarten through 3rd grade students. Students in each grade were stratified according to their race/ethnicity and oral language abilities (above average, average, and below average), with 80 to 138 students in each grade in the above average language group and 98 to 126 students from each grade in the below average language group. Each student produced eight oral language samples (four expository and four narrative). In phase 3, researchers analyzed the language samples using corpus linguistic analytical tools to produce vocabulary lists and grammatical descriptions for grades and subgroups. Samples were also scored using the Systematic Analysis of Language Software (SALT) for productivity indices (e.g., sample length, mean length of utterance in words, number of different words) and the Narrative/Expository Language Measures (NLM/ELM) Flowcharts for discourse and sentence complexity. In phase 4, researchers disseminated their findings through conferences and publications.
Control condition
Due to the nature of the research design, there was no control condition.
Key measures
Students were screened for above- or below-average oral language abilities using subscales from the Woodcock-Johnson IV Tests of Oral Language. The researchers coded language samples for the full range of grammatical and lexico-gramamtical features in English. These included grammatical features such as word classes, function word classes, stance features, and features reflecting grammatical complexity such as passive voice verbs and pre-modifying nouns. Using corpus linguistics, the researchers also produced vocabulary lists for grades and subgroups generated from the oral language samples. The researchers supplemented the corpus analyses with additional measurement methods to include productivity indices such as total number of words, number of different words, mean length of utterance, and to include content measures such as the completeness and complexity of story elements, information units, and complex linguistic features (e.g., subordinate clauses).
Data analytic strategy
The researchers used a corpus-based keyness' analysis to compare the frequency of vocabulary words in a language sample to the frequency of the same words in a reference text. They compared the language samples from the below-average students to the above-average student samples. In addition, they used analysis of variance (ANOVA) to compare the two groups on grammatical features. In terms of the supplemental analyses, a set of ANOVAs were conducted to compare discourse type and elicitation task on productivity of sample. Finally, based on age, gender, grade, mother's education, and ethnicity, 50 typically developing (TD) students were matched to a subsample of students with disabilities using propensity score matching. Multivariate analyses of variance (MANOVAs) were conducted to compare the various language indices in terms of their predictive ability.
Key outcomes
The main findings of this project are as follows:
People and institutions involved
IES program contact(s)
Products and publications
ERIC Citations: Find available citations in ERIC for this award here.
Publicly Available Data: Spencer, T.D. (2023). Academic Language of Primary Students (ALPS). https://nyu.databrary.org/volume/1632
Project Website: http://trinastoolbox.com/research_ALPS.html
Select Publications:
Almubark, N., Silva-Maceda, G., Foster, M. E., & Spencer, T. D. (2023). Indices of narrative language associated with disability. Children, 10 (11), 1815-1836. doi.org/10.3390/children10111815
Spencer, T. D. (2023, Jan). Is the reading crisis associated with an academic language crisis? Open Access Government, 288-289.
Spencer, T. D. (2023, Jan). Oral storytelling is important for reading, writing, and social wellbeing. Open Access Government, 286-287.
Spencer, T. D., Tolentino, T. J., & Foster, M. E. (2023). Impact of discourse type and elicitation task on language sampling outcomes. American Journal of Speech-Language Pathology, 32 (6), 2827-2845. doi.org/10.1044/2023_AJSLP-22-00365
Project website:
Supplemental information
Co-Principal Investigator: Reppen, Randi
- The context in which language samples are collected determine how much young children talk and the complexity of the language (Spencer et al., 2023). As opposed to narrative generation and expository retell contexts, expository generation and narrative retell conditions yield the largest samples with the greatest lexical diversity and syntactic complexity.
- Of an array of narrative language indices (e.g., number of different words, sentence complexity, story content), only narrative discourse consistently predicts language disability among K–3 students (Almubark et al., 2023).
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
