Study Habits and AI Coach
How to cite
Khoury, Stephanie El; Aucejo, Esteban; Peyre, Agustina Affonso; Vega, Martin Carbajal. 2025. "Study Habits and AI Coach." AEA RCT Registry. AEARCTR-0016486. https://doi.org/10.1257/rct.16486-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This study evaluates whether customized feedback and AI-based academic coaching can improve college students' study habits and academic outcomes. We conduct a randomized controlled trial with undergraduate students enrolled in intermediate economics courses at Arizona State University. Participants first complete a baseline survey measuring study strategies, time management, test anxiety, and AI familiarity. Students who consent to participate are randomly assigned to one of three groups: (1) a control group, (2) a treatment group receiving personalized written feedback on their study habits and , or (3) a treatment group receiving the same written feedback as in treatment 2, plus access to a custom-designed AI chatbot.
The feedback on their study habits includes scores, calculated from their responses to the baseline survey, along with a brief interpretation of that score.The chatbot offers tailored academic support by interpreting survey scores on study habits, answering freeform study-related questions, and helping students develop personalized study plans. Follow-up data will be collected at the end of the semester through a second survey and matched to academic records to assess the intervention's effects on beliefs about study habits, study habits strategies, and course performance.
Intervention
Participants and Setting:
- Undergraduate students enrolled in intermediate economics courses at Arizona State University (ASU).
- Recruitment begins after the first course exam via Canvas announcements.
- Baseline survey administered through Qualtrics.
Randomization and Stratification:
- Only students who provide informed consent are included.
- Stratified random assignment based on average high school GPA, implemented within Qualtrics.
- Randomly assigned to one of three groups: Control group, Personalized feedback group, and Personalized feedback + AI chatbot group
Survey & Video Procedures:
- After completing the baseline survey, students are shown a short video:
-- Chatbot group: video introduces the chatbot and explains how to use it.
-- Control and feedback groups: view a neutral, non-academic video (e.g., ASU campus content).
Post-Survey Emails (Immediately Following Survey Completion):
- Control group: receives a thank-you email.
- Feedback group: receives an email with personalized scores and written feedback on study habits.
- Chatbot group: receives the same personalized feedback email, plus: Access to the AI chatbot (hosted on ASU’s CreateAI Builder platform) and a usage guide explaining how to interact with the bot.
Chatbot Features:
- Tailored academic support based on individual survey scores.
- Generates study plans, answers study-related questions.
- Accepts course materials (e.g., lecture slides), which are processed only within-session.
Ongoing Intervention & Reminders (Before Subsequent Exams):
- Chatbot group: receives reminders on their study habit scores and encouraging bot use before each exam.
- Feedback group: receives reminders about their study habit scores before each exam..
- Control group: receives neutral study reminders before each exam..
- Reminders are aligned with exam dates and continue through the semester.
Follow-Up Survey (End of Semester):
- All students are invited to complete a follow-up Qualtrics survey
- Assesses:
-- Changes in study habits and beliefs
-- Intervention usage and perceptions (for treatment groups)
- Participation is voluntary; students may receive extra credit from instructors.
Experimental design
This project is a randomized controlled trial evaluating the effect of personalized feedback and an AI-powered chatbot on college students’ study behaviors and academic outcomes. The study population consists of students enrolled in intermediate economics courses at Arizona State University. After completing a baseline survey, consenting students are stratified by average high school GPA and randomly assigned at the individual level to one of three groups: (1) a control group, (2) a group receiving personalized written feedback on their study habits, or (3) a group receiving both personalized feedback and access to an AI chatbot. The chatbot offers academic coaching through interactive conversations and tailored study planning tools. Follow-up data are collected at the end of the semester via survey and matched to administrative academic records (with consent) to assess the effects of each intervention.
Design details
Cited by
No citation data available for this registration yet.