Project Activities
CELaRAI will carry out three main project activities over five years: (1) researching and developing the AIRE tool through an exploratory study, user-centered formative studies, and a pilot study; (2) addressing challenges related to responsible AI in schools and creating ethical guidelines for AI development and deployment; and (3) providing national leadership in the field of responsible AI for early literacy through research dissemination and capacity-building initiatives. To achieve CELaRAI's goals, an interdisciplinary team of researchers and expert advisors will collaborate across four thrusts: AI, Early Literacy, AI Ethics, and Learning Sciences and Methodology.
Focused program of research
The program of research focuses on developing and refining the AIRE tool for improving early literacy outcomes. First, the research team will develop a more fine-grained understanding about independent reading practices and materials in K–2 classrooms and conduct user-centered studies to understand teachers' AI use and perspectives to inform AIRE tool development. During the subsequent tool design phase, the research team will develop a prototype of AIRE by leveraging the best AI technologies identified, including large-language model based generative AI, conversational agents, and children's spoken language understanding. In addition, the research team will conduct formative user studies that will continue through the succeeding project years. When AIRE is fully developed, the research team will engage in a pilot study to test the effectiveness of AIRE for improving K–2 students' literacy outcomes, as well as AIRE's usability, feasibility, and fidelity to drive the iterative improvement of AIRE. Across all years of the project, the research team will examine challenges in responsible AI in schools and develop ethical guidelines for AI development and deployment.
National leadership and outreach activities
National leadership and capacity building activities will include creating several event series (such as annual conferences, online talk series, and practitioner forums), faculty and student training workshops, and activation of professional networks in both AI and literacy research. Additionally, the team will provide national leadership for responsible AI for early literacy research and practices through academic and public-facing dissemination.
Structured Abstract
Setting
This project will take place in urban, suburban, and rural public elementary schools in New York, Eastern North Carolina, and Michigan that serve culturally and linguistically diverse students from low- to mid-socioeconomic-status families.
Sample
In the exploratory study, there will be a national and local sample of 551 teacher participants and 54 children across grades K-2. During the tool design, each of approximately five iterations of user testing will engage 3-5 teachers and 10-20 children from our partner school districts. In the pilot study, participants will include 48 teachers and 720 K-2 students in eight schools in culturally and linguistically diverse and low- and mid-socioeconomic-status communities.
Research design and methods
In the exploratory study, the research team will conduct (1) a national teacher survey, (2) local teacher focus groups, and (3) local classroom observations and interviews with teachers and students. In the tool design phase, user-studies will involve think-alouds and laddered interviews of adults and children using various prototypes of AIRE tools. In the pilot study, the team will evaluate AIRE's (1) impact on early literacy outcomes through a school-year-long trial, using a multi-site cluster randomized experiment design, and (2) useability, feasibility, and fidelity, using mixed methods.
Key measures
The exploratory study outcomes include data regarding current classroom practices and materials for independent reading, and teachers' uses and perceptions of AI to inform tool design. Tool design outcomes include a refined user interface, and usability data identifying suitability for relevant audiences. The pilot study outcomes include early literacy learning outcomes measured by (1) Test of Phonological Awareness-2+ (grades K–1), (2) Test of Word Reading Efficiency-2 (grades K–2), (3) Test of Silent Reading Efficiency and Comprehension (grades 1–2), and (4) DIBELS-8th Ed. Oral Reading Fluency Benchmark Assessment (grades 1–2). The responsible AI outcomes (Years 1–5) include the development and provision of annual ethics training and guidelines for AIRE's development and deployment in schools. Capacity and national leadership outcomes (Years 1–5) include a website, a media platform, and professional networks for disseminating responsible AI in early literacy research and practices and supporting the wide adoption of AIRE.
Data analytic strategy
In the exploratory study, the research team will quantitatively and qualitatively analyze data from (1) the national teacher survey, (2) local teacher focus groups, and (3) local teaching observations and interviews with each observed teacher and six of their students. In the tool design phase, the team will use user-study data to identify challenges and opportunities for improvement of the usability of the tool. In the pilot study, the team will evaluate AIRE's (1) impact on early literacy outcomes using Hierarchical Linear Modeling (HLM) and (2) useability, feasibility, and fidelity using mixed methods.
Cost analysis strategy
The research team will apply the ingredients method to track: (1) equipment costs, (2) school support, and (3) teacher costs. They will use the CostOut tool for cost analysis and sensitivity analysis. Costs will be reported from the social perspective at these levels: (a) total implementation costs, (b) incremental costs compared to the control condition, and (c) costs per school, per classroom, and per student.
People and institutions involved
IES program contact(s)
Project contributors
Products and publications
Products: Products include (1) a fully developed AIRE system and evidence of its feasibility, usability, and efficacy to improve early literacy outcomes for diverse groups of K-2 students; (2) ethical guidelines for responsible AI in early literacy; and (3) a CELaRAI website and a media platform as well as professional networks of researchers and partners for disseminating responsible AI in early literacy research and practices and supporting a wide adoption of AIRE.
ERIC Citations: Find available citations in ERIC for this award here.
Related projects
Supplemental information
Co-Principal Investigators: Xiong, Jinjun; Christ, Tanya; Koyejo, Sanmi; Hoadley, Christopher; Alwan, Abeer; Hakkani Tur, Dilek; Strong, John; Tortorelli, Laura; Lee, Jaekyung
Partner Institutions: Buffalo Public Schools (Buffalo, NY); Erie 1 Board of Cooperative Educational Services (BOCES) (Erie County, NY); Avondale School District (Auburn Hills, MI); Genesee Intermediate School District (Flint, MI); Ingham Intermediate School District (Mason, MI); Michigan Association of Intermediate School Administrators (MI)
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
