An Investigation of Direct Instruction Spoken English (DISE) on Newcomers in Middle School: Examining Differential Response and Identifying Randomized Controlled Trial Implementation Barriers and Facilitators

In this presentation we will describe initial findings from a longitudinal study to examine the impact of Direct Instruction Spoken English (DISE) on the English oral language skills of middle school students who have recently immigrated to the United States. We study the impact of DISE using a design in which schools are randomly assigned to use DISE or their existing English language acquisition program during English as a Second Language instruction. The DISE curriculum seeks to improve oral language through a focus on explicit instruction in vocabulary, syntax, pronunciation, and morphology with an emphasis on repeated practice, corrective feedback and review. In addition to research supporting the effectiveness of direct instruction for English monolinguals, explicit, direct instruction for ELs is recommended across a range of policy reports and research recommendations (Carlo et al., 2004; Council of the Great City Schools, 2008; Francis et al., 2006; IES Practice Guide, Pashler et al., 2007; NYC Department of Education, 2014; Rivera et al., 2010).

The study uses a cluster randomized design. Twenty nine middle schools have participated in waves over the course of four years. The participating schools serve students in grades 6 through 8 from linguistically diverse backgrounds. Teachers in schools assigned to the DISE condition receive training at the beginning of their participation in DISE and follow-up coaching throughout the study. Fidelity of implementation is assessed in two ways. First, DISE coaches observe teachers twice each year. Second, the research team codes audio tapes obtained 3 times each year in both treatment and control classrooms using the Classroom Observations of Student-Teacher Interactions (COSTI) protocol. COSTI assesses the quality of instructional delivery through documentation of rates of teacher demonstrations, opportunities for independent practice, and corrective feedback (Smolkowski & Gunn, 2012; Doabler et al., 2015). Teachers in both treatment and control classrooms complete an implementation log to provide dosage information by gathering information on curricular activities, number of lessons completed, number of days of instruction, and any special considerations that affected instruction.

Researchers assess students in sixth through eighth grades on nonverbal ability and English oral language at the fall of each year to provide information on baseline skills. Researchers gather standardized measures of oral language and academic knowledge in the spring of each year for each cohort and one year following completion of the study for students who started with the study in sixth or seventh grade.

In this presentation will share preliminary findings from a differential response analysis from the first three years of four year study. Facilitators and barriers to implementation of an RCT across multiple states will also be discussed. Finally, the What Works Clearinghouse standards for “meets without reservations” will be discussed in the context of research with highly mobile populations.