Project Website: www.kooapps.com

Video Demonstration of the Phase I Prototype: https://youtu.be/_lxKe5Xvvhk

Purpose: In this project, the team will fully develop and test an artificial intelligence (machine learning) application that personalizes vocabulary learning for English Learners through visual game-based puzzles.  English Learners are the fastest growing segment of the K–12 public school population. According to the National Center for Education Statistics, the percentage of students who were ELs in 2017 was 10.1%, or 5 million students. By 2025, an estimated one in four K–12 students will be English Learners.

Project Activities: In prior research and development the team developed Pictoword, an application that uses visual, puzzle, and game-based mechanics for players to build fun vocabulary words by combining pictures. In 2020 in Phase I, the team developed a new prototype to personalize the learning experience for individual students who are English learners to build their English vocabulary. The prototype included an artificial intelligence engine, enhanced educational content, and a teacher dashboard to track student performance. At the end of Phase I, researchers conducted a classroom feasibility study with four grade 3 teachers and 42 English Learners. The prototype operated as intended as all students were able to complete puzzles to build words, and teachers were able integrate activities into classroom practice and review results from student progress on a reporting dashboard. Results also demonstrated that students increased from pre-test to post-test on researcher developed measure of vocabulary and spelling.

In Phase II, the team will expand content, create new puzzle mechanics, introduce retention features, improve the personalization algorithm, and broaden the platform availability, and improve reporting and content customization features to the teacher dashboard. After development is complete, the researchers will complete a pilot study to examine the usability and feasibility, fidelity of implementation, and promise of the app to increase students vocabulary. The study will include 24 grade three classes 20 students per class and approximately 480 students overall. Half of the classes will be randomly assigned to use Kooapps and the other half will complete business-as-usual activities. The study will compare student scores on standardized pre- and post- learning outcomes for vocabulary acquisition. The team will gather cost information using the "ingredients method" and will include all expenditures on things such as personnel, facilities, equipment, materials, and training.

Product: Pictoword School is an application for personal devices for English Learners that presents a set of fun and engaging word puzzles to support vocabulary acquisition, including word recognition, spelling,pronunciation, and use in a sentence.  The intervention will include two versions. In Classroom Mode groups of students will engage in teacher-led lessons and in Freeplay Mode students will solve puzzles on their own during free time in school or at home. An artificial intelligence (machine learning) engine will personalize the learning experience by determining the level for each player, and present more challenging or easier puzzles at students' level of vocabulary acquisition and adjusts as they go. A website will include a dashboard to provide teachers with formative assessment results to inform practice, an interactive dictionary, and professional development modules on how to integrate the app into their classroom practice and create custom puzzle packs aligned to standards for course in grades K to 6.

Related IES Project: Pictoword School: Combining AI (Machine Learning) and Game-Based Learning to Support English Learners (91990020C0069)