Academic Language of Primary Students (ALPS)

The ALPS project was funded by the Institute of Education Sciences (IES), U.S. Department of Education. I worked with expert corpus linguists from Northern Arizona University (Randi Reppen and Douglas Biber).

Academic language is specialized in its form and function. It consists of word-, sentence-, and discourse-level patterns in oral and written language that are more prevalent in academic contexts compared to social contexts (e.g., conversation). Understanding and use of academic language are critical for being successful in school. In fact, many call it the language of schooling needed to acquire and express knowledge. Even though most educators acknowledge that language is important for reading and writing, little is known about the dimensions of academic language and how to teach it, especially in primary grades.

Want to learn more about academic language? Download grade-level profiles here
KINDERGARTEN PROFILE 1ST GRADE PROFILE 2ND GRADE PROFILE 3RD GRADE PROFILE

To learn about the academic language that young students produce, we collected over 9,000 oral and written language samples from 2018-2020. More than 1000 K-3 students contributed up to nine language samples each. Using a strategic sampling procedure, care was taken to ensure students were culturally and racially diverse and that demographic groups were represented at all language levels (i.e., above average, average, and below average).

We elicited oral and written language samples in multiple contexts, including narrative and expository discourse types and using retell or generation tasks. By creating a set of standardized materials and procedures, we could investigate the influence of elicitation context and task on students’ language and be able to make direct comparisons (e.g., narrative vs. expository, retell vs. generation).

The ALPS elicitation materials consist of nine photo sets for narrative language sampling and nine photo sets for expository language sampling. For the narrative photo sets, we ensured that diverse children were depicted in the photos. When they were unavailable to purchase, we hired a photography and created our own images with African American and Latino children as characters. The procedures for both narrative and expository conditions included allowing the students to choose which photo set they wanted to talk about. When we field tested the procedures, we found that African American students and Latino students selected photo sets in which the children matched their own race or ethnicity. By letting students choose which set of photos they talk about, we could reduce unwanted bias in the materials.

The oral and written narrative and expository language samples have been transcribed and using multiple methods. First, we used corpus linguistic analysis tools to examine the relative frequency of specific words and grammatical features (Thanks to Randi Reppen and Doug Biber from Northern Arizona University!).Second, we analyzed all the samples using Systematic Analysis of Language Transcripts (SALT), which yielded information about the number of total words (NTW), number of different words (NDW), and mean length of utterance in words (MLU-W) of each sample. Additionally, our well-trained research team coded the SALT transcripts for subordination and grammaticality. Third, the research team also scored each narrative and expository language sample using the Narrative Language Measures (NLM) Flowchart or Expository Language Measures (ELM) Flowchart, which are part of the CUBED-3.

Using the Flowcharts, researchers scored each sample for Discourse Complexity and Sentence Complexity. The NLM Flowchart has been used in a handful of previous research (Kirby et al., 2021; Petersen et al., 2020; Petersen et al., 2022; Spencer et al., 2013; Spencer & Petersen, 2018) but the ELM Flowchart was new. During this project, we examined the psychometric properties of both and made iterative revisions based on our ongoing analyses. The NLM Flowchart has been refined and the psychometric evidence is adequate to recommend its broad use. However, the ELM Flowchart requires some additional work. The Sentence Complexity section of the ELM Flowchart works just fine, but we are still examining the Discourse Complexity section. Also, there is a section on each flowchart for which we have not yet completed the analyses—Writing Conventions. We have collected an enormous amount of data so that we can examine the psychometric properties of the Writing Conventions section, but we have not yet completed it. So, if you choose to use the Writing Conventions section, please do so with caution.

We developed the Flowcharts because practitioners (e.g., teachers and SLPs) are very busy and need easy to use assessment tools that can inform intervention and instruction. Although it can be useful to transcribe language samples prior to scoring using the Flowcharts, it is time consuming. When scoring using the Flowcharts, transcription is not required. Practitioners can listen to an audio recording of an oral language sample and score as they listen. Importantly, the Flowcharts have a direct link to intervention because if a student consistently produces language samples without a specific feature, that feature should be explicitly targeted during intervention. An advantage of the Flowcharts for use in primary grades is that they can be used on both oral and written language. They can be used as a screening tool for identifying students with Developmental Language Disorder (DLD) or to monitor progress over time. They are particularly helpful to teachers who need a reliable method for scoring students’ writing.

Even though our empirical papers are not yet published, we have created academic language profiles for each grade. They have useful guidance for educators about teaching academic. Please check back periodically for links to our ALPS publications. In the meantime, download the ALPS elicitation materials and procedures for free and visit https://www.languagedynamicsgroup.com/cubed/ to get the Flowcharts free with the CUBED-3 suite of assessments.