Improving STEM Outcomes for Young Children With Language Learning Disabilities

Improving STEM Outcomes for Young Children With Language Learning Disabilities by Intervening at the Intersection of Language and Scientific Thought.

The sophisticated language of science can be a barrier to Science, Technology, Engineering, and Math (STEM) learning, especially for children who have specific language impairment (SLI). The purpose of this randomized controlled trial is to test vocabulary and grammar interventions embedded in a small-group inquiry-based science instruction for their potential to ameliorate language deficits that impede science learning. Participants will be 54 preschoolers or kindergartners with SLI. Proximal and distal probes will reveal their mastery of taught and generalized language and science concepts.

In this study the investigators focus on a subset of at-risk students who find the language of science to be a barrier to the learning of science. These are the nearly 3 million children in the U.S. who have a learning disability called specific language impairment (SLI). Children with SLI present with deficits in spoken grammar and vocabulary and they are 3.9 to 8.1 times more likely to have reading deficits than children in the general population. Specific Aim #1: To determine whether science-relevant language intervention enhances the learning of science concepts in young children who have SLI. Specific Aim #2: To determine whether science-relevant language intervention facilitates generalization of science concepts and practices in young children who have SLI. Fifty-four 4-to-7-year-olds who have not yet begun 1st grade, who are monolingual speakers of English, and who have SLI will participate. The investigators will adopt a Randomized Controlled Trial design, randomly assigning participants into three intervention conditions: science + phonological awareness practice (the control arm), science + vocabulary supports, and science + grammar supports, followed by a brief withdrawal phase in which all three groups receive science only instruction. Pre- and post-measures will reveal the extent of learning in each condition and comparisons between conditions will reveal whether the grammar and vocabulary supports improved learning. The hypothesis is that the language and learning of science are integrally related. Therefore, the investigators will use evidenced-based language interventions to improve the children's science-relevant language skills, with the prediction that this will cascade into changes in the acquisition of science concepts and practices: Children in the science + language intervention conditions will show greater gains in taught science concepts after the 4-week intervention period than children in the control arm. The benefit of the science + language interventions will remain after the language supports are withdrawn, that is, children in the science + language intervention conditions will show greater gains in taught science concepts during the withdrawal week than children in the control arm. Children in the science + language intervention conditions will show greater gains from pretest to posttest on measures of generalized science concepts and practice than children in the control arm. Children who demonstrate the greatest improvement in the use of the language targets will also demonstrate the greatest improvements in taught concepts, generalized concepts, and generalized practice knowledge. Children will benefit from language supports directed at vocabulary as well as those directed at grammar, but these supports may differently benefit the science learning process. The first step is to document that the language supported interventions resulted in improved language abilities by comparing performance on probes of grammar and vocabulary at posttest to pretest performance. The expectations are significant changes in vocabulary knowledge for the vocabulary intervention condition as compared to the other two conditions, and significant changes in use of complement clauses for the grammar intervention condition as compared to the other two conditions. The next step is to test the predictions associated with the specific aims via a series of binomial mixed models. Mixed models are appropriate for designs with unbalanced cell sizes due to missing data (due to non-response and dropout). There will be one model for targeted science concept outcomes with condition (control arm, science + vocabulary, science + grammar), language support (present, withdrawn), and condition x language support as the independent variables (Predictions 1 and 2). If data plotting suggests that effects are specific to the type of concepts being taught (e.g., physical science vs biological science), then we will build a second model to explore differences related to concept type. There will also be one model each for generalized concepts and generalized practice outcomes with condition (control arm, science + vocabulary, science + grammar) and time (pretest and posttest) as independent variables (Prediction 3). Within-subject correlation will be accounted for with random subject effects. Additional random effects will be determined by selecting the model with the best model fit (lowest AIC value). In each of the three models, it is further expected that amount of improvement in grammar and vocabulary are moderators between the outcome and the other factors (Prediction 4). To assess this prediction, performance on the language tests will be included as covariates. The expectation is that performance on the language probes after instruction will be a significant predictor of science learning, and that including performance on the language probes as a covariate will eliminate the effect of condition because language performance will be the main factor predicting science performance. These models also allow comparison of the effectiveness of the grammar- and vocabulary-supported conditions (Prediction 5).

Language Disorders, Communication Disorders, Neurobehavioral Manifestations, Developmental Language Disorder, Randomized Controlled Trial

Although it is impossible to mask the assessors of the proximal outcome measures, those assessing some of the distal outcomes can be masked. Specifically, the people scoring the results of the Lens on Science Assessment (a distal measure of generalized scientific content and process) will be masked to the treatment arm assignments. Also, the people transcribing, coding, and analyzing the words and gestures used in the science story retell task (a distal measure of generalized scientific practice outcomes) will be masked to the treatment arm assignments.

In all conditions, science is taught via the Full Option Science System Next Generation Edition (FOSS, 2015, https://www.fossweb.com/) curriculum that involves 1) Prediction, 2) Experiment, 3) Journal/Reflection, and 4) dialogic reading centered around a given theme such as plant life. In the control condition, a minimum of six phoneme identifications and five rhymes will be incorporated into each lesson of this curriculum. While these activities are likely to improve the children's awareness of the sounds of the language (a foundational skill for learning to read), they are not likely to improve their access to the science being taught. Therefore, this intervention constitutes a placebo.

In the science + grammar condition, focused stimulation, an intervention commonly used to target expressive language, will be used to treat complement clauses during the FOSS activities. The approach is incidental, rather than explicit. The active ingredients are models and recasts of the target structure. Recasts occur when an examiner responds to a child's naturally occurring utterance by expanding or extending the child's utterance to include a target grammatical structure. Recasts and/or models will be provided at an average rate of one per minute, an accepted therapeutic dose.

This arm involves Robust Vocabulary Instruction, an explicit approach that emphasizes multiple and rich encounters in authentic contexts to promote depth of semantic knowledge of 20 words that pertain to scientific practices applicable to the FOSS lessons. The children receive a cumulative exposure of at least 20 times per word (a minimum of 5 times per each of four lessons) and at least 4 chances to produce the word (a minimum of 1 chance per each of four lessons).

The goal is to facilitate phonological awareness during science lessons. Because phonological awareness does not relate to science learning, this is a placebo.

The goal is to facilitate science learning by providing models, prompts, and experiences with grammatical structures related to science processes.

The goal is to facilitate science learning by providing models, prompts, and experiences with vocabulary related to science processes.

The outcome relevant to aim 1 will be measured with eight 10-item, proximal concept assessments (to cover content taught during the intervention period) and two 10-item proximal assessments (to cover content taught after the intervention has been withdrawn). These will all be administered during the withdrawal week. These will be adapted from the FOSS I-check probes (e.g., Item 1: What is in the night sky? Stars? Sun? Moon?). From these, an accuracy score will result. The investigators will be able to determine whether the children learn more of the target science concepts in the language-supported conditions than in the control arm. Comparison of the performance on these assessments will allow determination of whether any improved access to the curriculum remains after the language supports are withdrawn.

For science concepts and processes, the distal measure is the Lens on Science Assessment (https://ies.ed.gov/ncer/projects/grant.asp?ProgID=7&grantid=805). A score reflecting mastery of science concepts will result. This will be administered to children in all three conditions in the three weeks before intervention and again in the three weeks after withdrawing the language supports in the active conditions.

For science practices, the distal measure is the retelling of a Sid the Science Kid episode (http://pbskids.org/sid/). A score reflecting completeness of the retelling will result. This measure will be administered to children in all three conditions in the three weeks before intervention and again immediately after withdrawing the language supports in the active conditions.

An investigator-created probe involving 20 complement clause elicitations will be administered to children in all three conditions in the three weeks before intervention and again in the three weeks after the withdrawal week. A score reflecting accuracy of production will result.

An investigator-created probe involving 20 vocabulary definition and levels of knowing items will be administered to children in all three conditions in the three weeks before intervention and again in the three weeks after the withdrawal week. A score reflecting depth of knowledge will result.

The rhyme detection-phoneme matching probes from Gillon (2005) http://www.canterbury.ac.nz/media/documents/education-and-health/gail-gillon---phonological-awareness-resources/assessment/PA-Assessment-probe-instructions-(1).pdf will be administered to children in all three conditions in the three weeks before intervention and again in the three weeks after the withdrawal week. A score reflecting accuracy of phonological awareness will result.

Inclusion Criteria: Age between 4 and 7 years Not yet begun first grade Speaks English as their primary language Has SLI confirmed by 1) a standard score of 94 or lower on the Structured Photographic Expressive Language Test, 3rd edition (SPELT-III, Dawson, Stout, & Eyer, 2003) OR below a scaled score of 7 on the Diagnostic Evaluation of Language Variance™-Norm Referenced (DELV-NR, Seymour, Roeper, & de Villiers, 2005) syntax subtest; AND 2) performing below age-relevant cutoffs on the Dollaghan and Campbell (1998) Nonword Repetition Task OR enrollment on a clinical caseload. Nonverbal matrices t score of 35 or higher on the Developmental Abilities Scale Passes a pure-tone audiometric screening administered according to the standards of the American Speech-Language-Hearing Association (ASHA, 1997) Can produce simple sentences that contain a subject and a verb. Performs with less than 40% accuracy on expressive probes of complement clauses prior to study onset Performs with less than 40% accuracy on vocabulary definition probes prior to study onset Exclusion Criteria: Other diagnosed neurodevelopmental disorders (e.g., autism, Down syndrome) via parent report or significant sensory or motor impairments (e.g., severe vision impairment uncorrectable by glasses) Exposure to a language other than English at home or school more than 20% of the time.

Dawson, J., Stout, C., & Eyer, J. (2003). Structured Photographic Expressive Language Test, 3rd edition. DeKalb, IL: Janelle Publications.

Dollaghan C, Campbell TF. Nonword repetition and child language impairment. J Speech Lang Hear Res. 1998 Oct;41(5):1136-46. doi: 10.1044/jslhr.4105.1136.

Seymour, H. N., Roeper, T., & de Villiers, J. G. (2005). DELV-NR (diagnostic evaluation of language variation) norm-referenced test. San Antonio, TX: The Psychological Corporation.