SBIR-STTR Award

This project will develop a prototype of an online program to support students in middle and high school who are English learners (ELs) in reading passages in English. The prototype will present passages as diglot weave texts. These texts are created by blending English and Spanish words together, with passages continually changing as readers gain skill in the target language. The prototype will include a machine learning engine that automatically translates single words or small phrases into a target language to support vocabulary acquisition and improved reading comprehension. At the end of Phase 1, in a research study with 100 middle school EL students, researchers will determine whether the prototype functions as intended, whether students are engaged when reading passages, whether the passages adjust to the reading level of the individual students, and whether the prototype shows promise for supporting the acquisition of the English language.

Purpose: In this project, the team will fully develop and test an online program to support students in middle and high school who are English Learners (ELs) in reading passages in English. The product will present diglot weave passages that blend English and Spanish words together, with a machine learning engine continually adjusting the words as readers gain skill in the target language. English Learners consistently receive lower scores on standardized tests than their non-EL peers; especially adolescent ELs in middle and high school. Reports demonstrate that 8th grade ELs score lower than their non-EL peers in every educational category and are considered "below basic" over 60% of the time across all academic categories. Project Activities: During Phase I in 2020, the team developed a prototype of an online program for middle and high school students who are English Learners (ELs) that presents diglot weave passages that blend English and Spanish words together, with passages continually changing as readers gain skill in the target language. The prototype includes a machine learning engine that automatically translates single words or small phrases into a target language to support vocabulary acquisition and improved reading comprehension. At the end of Phase I, researchers completed a pilot study with 13 middle school students who are English Learners and four teachers. The prototype operated as intended when students read passages and students could use the platform to practice and learn vocabulary. Research also demonstrated that students exhibited positive gains on their self-perceived knowledge of vocab words, showing nearly 20% improvement from pre- to post- test on a researcher developed measure. In Phase II, the team will fully develop the mobile app with a browser extension, a machine learning algorithm to handle more speech and word phrases, and the functionality of educator dashboards. After development is complete, the researchers will complete a pilot study to examine the product's usability and feasibility, fidelity of implementation, and promise to increase vocabulary acquisition, reading comprehension, and attitudes about reading compared to regular instructional practices. The study will include 20 middle schools with higher proportions of students who are English Learners. Students who are English Learners in half of the schools will be randomly assigned to use Loomvue and English Learners in the other half of schools will complete business-as-usual activities. Pre- to post- scores will be examine using standardized measures including the Gates-MacGinitie Reading Tests. The researchers will gather cost information using the "ingredients method" and will include all expenditures on things such as personnel, facilities, equipment, materials, and training. Product: Loomvue will be an online program to support students in middle and high school who are English learners in reading passages in English. Researchers will present the product as what linguists refer to as "diglot weave" passages that blend English and Spanish words together, with a machine learning engine continually flipping or adjusting the words from one language to the other as readers gain skill in the target language. A Brower extension-based plug-in application will be available to be used with any website content. A website will include a dashboard to provide teachers with formative assessment results to inform practice on student vocabulary acquisition and reading comprehension, and professional development modules on how to integrate the intervention into their classroom practice.