|Title:||Generating Large and Sustained Impacts on Early Language Skills: Evaluation of Timing and Duration of Intervention|
|Principal Investigator:||Lonigan, Christopher||Awardee:||Florida State University|
|Program:||Cognition and Student Learning [Program Details]|
|Award Period:||4 years (7/1/2016–6/30/2020)||Award Amount:||$3,300,000|
|Type:||Efficacy and Replication||Award Number:||R305A160241|
Co-Principal Investigator: Beth Phillips
Purpose: The purpose of this project is to examine the relative efficacy and sustained impacts of three variations in timing and duration of a language and vocabulary intervention for students in prekindergarten and kindergarten with below or low average language skills. Because language weaknesses place children at heightened risk for reading comprehension and general academic difficulties, an intensive early intervention approach is necessary. The intervention is an integrated combination of two interventions with prior evidence of efficacy: 1) Dialogic Reading-Enhanced (DR-E), which will be used to teach vocabulary to students and 2) Language in Motion (LIM; developed through a prior grant from the Reading for Understanding Initiative, Examining Effective Intervention Targets, Longitudinal Intensity, and Scaling Factors for Pre-K to 5th Grade Student Comprehension), which will be used to teach syntax.
Project Activities: In order to test the relative efficacy and sustained impacts of these two interventions, researchers will randomly assign students to one of four intervention conditions: (a) intervention during prekindergarten only; (b) intervention during kindergarten only; (c) intervention in both grades; and (d) business as usual instruction in both grades. The research team will administer the intervention and collect data across two cohorts of students, one beginning in Year 1 and one beginning in Year 2. The research team will follow the students through the end of first grade (during Years 3 and 4 of the grant). In Year 4, the research team will analyze the data and disseminate the findings.
Products: The products of this project include additional evidence of the efficacy of DR-E and LIM, a determination of the optimal timing and dosage of this combination of interventions, and peer reviewed publications.
Setting: Schools participating in this study are public schools and other prekindergarten settings in rural and suburban Florida and Georgia. These sites serve a disproportionate number of students from backgrounds of poverty.
Sample: Study participants will include prekindergarten students recruited and screened for eligibility in each of two consecutive school years within 33 prekindergarten sites, for a total of nearly 1,900 students screened and 1056 (estimated 55% of those screened) qualifying for the study. Eligibility criteria include: (1) performance at or below the 40th percentile on the Expressive One Word Picture Vocabulary Test –4th edition (EOWPVT); (2) no significant hearing or visual impairments that would impede participation in the assessment or intervention protocols; and (3) scores above a scale score of 4 (i.e., two standard deviations below the mean) on the Block Design subtest of the Wechsler Preschool and Primary Scales of Intelligence.
Intervention: The intervention includes an integrated combination of two previously validated language- and vocabulary-focused small group, Tier 2 interventions. DR-E exposes students to narrative and informational texts and provides instructional techniques to encourage students to use complex expressive language. LIM provides instruction on syntax through narrative stories and lesson activities. The combined intervention alternates between DR-E and LIM lessons.
Research Design and Methods: The research team will conduct a randomized controlled trial, using a student-level, four-group, intent-to-treat design that will evaluate different doses and sequencing for intervention. Researchers will screen students for eligibility and randomly assign eligible students to one of four intervention conditions: (a) intervention during prekindergarten; (b) intervention during kindergarten; (c) intervention in both grades; and (d) business as usual instruction in both grades. Researchers will recruit prekindergarten students in two cohorts, with one beginning in Year 1 and one beginning in Year 2. Participating teachers will receive professional development sessions and coaching, and will teach lessons four days per week for 18 weeks during the school year (equaling 30 hours per year). The research team will administer assessments to students in the fall and spring of prekindergarten and kindergarten as well as spring of first grade. In addition to these time points, the team will administer researcher-developed assessments mid-year during prekindergarten and kindergarten. The research team will conduct observations and collect audio recordings to measure fidelity of implementation.
Control Condition: The control condition is business-as-usual Tier 1 classroom instruction. The prekindergarten-only and kindergarten-only conditions also serve as counterfactuals to the intervention in the both-grades condition.
Key Measures: Primary measures include researcher-developed assessments of vocabulary and syntax and standardized measures of vocabulary and language, including the EOWPVT, the Syntax Construction subtest from the Comprehensive Assessment of Spoken Language, and the Oral Comprehension subtest of the Woodcock Johnson Test of Achievement-III, as well as the following computer-based Florida Center for Reading Research (FCRR) Reading Assessments: Vocabulary Pairs, Sentence Comprehension, and Following Directions. The research team will measure executive function skills with computer-based versions of the Word-Span Reversed and Head-Toes-Knees-Shoulders assessments. They will also use the EOWPVT and the Block Design subtest of the Wechsler Preschool and Primary Scales of Intelligence to screen for eligibility.
Data Analytic Strategy: The research team will use multi-level models that take into account the nesting of children in schools and cohort, with student at the level of randomization. All models will include four covariates: child age, initial score on the outcome measure, pretest score on the non-verbal cognitive ability measure, and initial score on the screening measure.