|Title:||WKI-Writing: Impact of Word Knowledge Instruction (WKI) on Writing Outcomes of 5th Grade Students|
|Principal Investigator:||Wood, Carla||Awardee:||Florida State University|
|Program:||Low-Cost, Short-Duration Evaluation of Education Interventions [Program Details]|
|Award Period:||2 years (07/16/2018 – 07/15/2020)||Award Amount:||$249,230|
|Type:||Other Goal||Award Number:||R305L180019|
Co-Principal Investigators: Timetia Creed (HCPS); Christopher Schatschneider (FSU); Linda Gaughan (HCPS)
Partner Institution: Hillsborough County Public Schools
Purpose: Writing achievement is an important aspect of language and literacy performance for ELs and students from low socioeconomic backgrounds. Writing and oral language skills are integrally intertwined, as writing relies on interconnected language skills including word knowledge. Academic language (referring to the language of schooling that occurs in content-area text) is thought to be central to language and literacy achievement. Academic language is particularly important for students from low-income households and English learner students who often lag behind in language and literacy achievement. The proposed study is designed to build upon past work to evaluate the effects of vocabulary instruction on writing outcomes of students as measured by classroom writing samples and a standardized state assessment measure.
Project Activities: The FSU team will take the lead in coordinating all research and dissemination-related activities and will work closely with HCPS to ensure that random assignment is implemented appropriately and that all required data are collected in a timely and efficient manner. The school district will take the lead in preparing de-identified data sets and providing them to the FSU team. The FSU team will then conduct statistical analyses and collaborate with co-investigators to disseminate findings and share lessons learned.
Products: The project will result in agency-wide debriefings, consumer-friendly written briefs, and a Toolkit for using WKI in elementary schools.
Setting: This study will take place in elementary schools in Hillsborough County Public Schools (HCPS), a large district in central Florida that serves an increasing number of low-income and English Learner students.
Population/Sample: Participants will include approximately 3,907 grade 5 students and 100–120 teachers from 44 high-poverty schools in HCPS.
Intervention: Word Knowledge Instruction (WKI) consists of daily 15-minute lessons taught by grade 5 ELA teachers over 20 weeks. The instruction entails explicit code-based instruction, focusing on three core components: a) morphological awareness; b) vocabulary; and c) discourse connectives. Evidence-based strategies are integrated in activities including: a) explanation of the meaning of the target base and affix; b) identification of the targets in oral and written passages; c) discussion of using morphological skills to deduce meaning; and d) small-group games; sentence construction, and completion tasks for active practice integrating written language responses.
Research Design and Methods: In this study, researchers will randomly assign teachers in HCPS to WKI and business-as-usual ELA instruction.
Control Condition: Teachers and student that are not involved with the intervention will continue with "business-as-usual ELA instruction."
Key Measures: Systematic coding of students’ written language data will include the number of different words, multi-morphemic words, and connectives. Additionally, researchers will use administrative data of teachers' scores of students writing for cohesiveness, content, and writing conventions using the FSA writing rubric on Fall and Spring written language samples.
Data Analytic Strategy: The research team will estimate main effects of intervention on students' writing performance via a series of three-level hierarchical linear models, with students nested within teachers nested within schools. The models will account for blocking of teachers within schools and will take a covariate adjusted approach to the estimation of change over time between treatment and control teachers. To address the first two research questions, researchers will estimate six separate models and to address the third and fourth research questions, researchers will add one student level predictor (English learner status) and the cross-level English learner status by Treatment interaction to the same six 3-level hierarchical linear models.