Study of AI and Human Tutoring for Early Reading Development

We are conducting a research study to understand how different types of reading support help children learn.

In this study, we are comparing:

  • An AI-powered reading tutor
  • One-on-one human tutoring
  • Typical classroom instruction

Amira Learning is partnering with researchers at Boston University through the Evidence-Based AI in Learning (EVAL) Initiative to carry out this study.

Our goal is to determine how well AI tutoring supports reading development, how it compares to human tutoring, and which approaches work best for different students.

This research will help schools make informed decisions about how to support children’s reading growth.

Who Can Participate

First and second graders who speak English and live in the U.S.

How the Study Works

This study will take place on video calls, live with a researcher! Clicking on the “Enroll here” button will send you to a brief survey to enroll your child in the study; a researcher will follow up within two weeks about the first meeting. In the first meeting, your child will play some word games to help us understand their reading level. You will also be asked to complete brief online questionnaires about your child.

Then, your child will be placed into one of three groups by chance (like flipping a coin). Your child will have an equal chance of being assigned to all study groups. Depending on the group your child is assigned to, they may:

  • Receive free access to Amira, an AI-based reading tutor
  • Work one-on-one with a human reading tutor
  • Participate in their regular day-to-day activities

These sessions will be scheduled with the study team.

After 10 weeks, your child will play some more word games to see how their reading improved. No matter which study group they were assigned, we will learn about all children’s reading and language knowledge; we will share results of children’s reading and language assessments at the conclusion of the study.

All children will receive access to Amira; after the final assessments are completed, children who are not assigned to the Amira group will be given free access to Amira.

Check eligibility

What We’re Studying

This study investigates what features of an AI-based reading tutor are most effective at improving student learning. It looks more specifically at Amira Learning, a popular AI-based reading tutor, and how it affects K-3 students’ literacy skills. While AI-based tools are increasingly used in education, scientists do not know which features are most important for children’s learning. Our study will help us understand which parts of Amira improve children’s literacy skills and how it compares to one-on-one human tutoring. Our findings could also be used by schools and policymakers to better choose AI-based tools and by companies to design better AI-based tools.

Duration

Participation begins with a 1-hour initial session to assess your child’s current literacy and language skills. This is followed by a 10-week intervention. During these 10 weeks, children will engage in three 30-minute sessions, totaling 1.5 hours of structured intervention time per week. After the intervention, children will complete a 1-hour final session to measure progress and learning over time.

Compensation

After your child completes the post-assessment games, we will send you a $10 Amazon gift card within a week and children will receive a certificate in recognition of their help in research and advancement of science. To be eligible for the gift card, (1) your child must fall within the study’s grade range, (2) you must provide a valid consent statement, and (3) your child must be present and visible during the sessions. Please note that each child may only participate once.

This study is conducted by Ola Ozernov-Palchik and funded by Boston University. Email us at evalcoll@bu.edu for questions or more information.