Project Activities
The researchers conducted an RCT replication study to investigate the effects of TELL on the language and early literacy skills of typically developing children from economically disadvantaged families. They recruited and randomly assign 58 preschool teachers to treatment (N=30) and control conditions (N=28). Teachers received professional development training and coaching to implement the curriculum with fidelity. The researchers collected data from parents and teachers, conducted classroom observations, and assessed three cohorts of children to evaluate impacts of the intervention on child outcomes. Because of school closures during the COVID-19 pandemic, they could not collect spring semester curriculum-based measures or end of year data from the third cohort of children.
Structured Abstract
Setting
This study took place in public preschool programs operated by school districts and agencies in the greater Phoenix metropolitan area that serve children from low socioeconomic status backgrounds, including community and Head Start preschool programs.
Sample
Study participants included 58 preschool teachers (30 TELL; 28 BAU) and 324 (174TELL; 150 BAU) racially and ethnically diverse children from low-income families.
Intervention
The Teaching Early Literacy and Language (TELL) curriculum is a tier 1, whole-class curriculum designed to increase both the quality and quantity of code-related and oral language instruction in preschool classrooms. TELL was developed with IES-funding. The curriculum package includes materials, scope and sequence, activities, and lesson plans. The package also includes professional development (coaching, professional learning communities, and formal training) to promote high fidelity of implementation. TELL includes 32 weeks of instruction during the school year. There are 14 thematic units, each two weeks in length. Review weeks occur every fifth week. Prior to the review week, curriculum-based measures are administered to measure children's progress on unit objectives.
Research design and methods
Researchers conducted the study over a four year period employing a three-cohort design in the first three years to complete a cluster randomized trial. Each TELL cohort received teacher training and coaching to implement TELL. Parent questionnaires, direct assessments of children and fidelity of implementation data were collected for both TELL and BAU conditions. During the spring of the third year of implementation schools closed due to COVID 19. This meant that not all planned spring curriculum-based measures or post tests were collected for the third cohort of children. In year 4, researchers conducted analyses to evaluate the efficacy of TELL and to disseminate research findings.
Control condition
Classrooms in the business as usual condition used a variety of instructional approaches. Some classrooms used a comprehensive curriculum with language and literacy teaching components and other classrooms used teacher selected activities mapped to the state early learning standards.
Key measures
Primary measures included standardized direct child assessments of children's early language and literacy skills, measures of fidelity of implementation fidelity (adherence and dosage) and the classroom language environment, and teacher questionnaires. Direct assessments included the Woodcock Johnson-IV Tests of Oral Language, and the alphabet knowledge, vocabulary/oral language comprehension, and phonological awareness subtests from the Preschool Early Literacy Indicators (PELI). TELL curriculum-based measures included upper case letter naming, letter sounds, print awareness, beginning sounds, narrative story recall, expository summary, receptive vocabulary, and expressive vocabulary. Classroom observation measures included the Classroom Assessment Scoring System (CLASS) and the Early Language and Literacy Classroom Observation-Revised (ELLCO-R).
Data analytic strategy
After testing for cohort effects and finding none, children from the Year 1, Year 2 and Year 3 cohorts were combined into a single sample for analytic purposes. All analyses included maternal education as a covariate. For most curriculum-based measures researchers used growth curve models (linear mixed models) that include fixed effects for time, time squared (quadratic trend model), treatment, treatment by linear time, and mother's education as a covariate. The model included random effects for the growth curve model (random intercept and linear slope) and a random effect for classroom teacher. A treatment by quadratic time trend interaction was investigated for all variables, but none were significant. Scores on three of the CBMs, Receptive Vocabulary, Expressive Vocabulary, and Expository Summary, were not repeated measures, so growth curve models were not applied to these CBMS. Instead, TELL was compared to BAU at each time point.
Key outcomes
The key outcomes of this replication study will be included here once they are published.
People and institutions involved
IES program contact(s)
Products and publications
Products: Teaching Early Literacy and Language (TELL) Preschool Curriculum; TELL Curriculum-Based Measures; TELL Professional Development Training
Project website: https://flounder-porcupine-47ep.squarespace.com/tell
ERIC Citations: Publications from this study are available here.
Project website:
Supplemental information
Co-Principal Investigators: Wilcox, M. Jeanne; Marley, Scott C.; Reiser, Mark
For the PELI, the mixed model included a random effect for classroom and a fixed effect for occasion instead of random intercept and slope terms for trend over time. The unstructured (UN) model was selected for R_i . Given the selected covariance structure, the model fixed effects were estimated by the method of estimated generalized least squares using SAS PROC MIXED software. For the Woodcock-Johnson IV Tests of Oral Language the mixed model for the Listening Comprehension and Phonetic Coding cluster score variables was the same as the model used for the PELI measures.
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
