College of Education
Virtual Learning Lab
Led by Walter Leite · Human Development and Organizational Studies in Education
What we study
The Leite Lab studies how virtual learning environments affect student learning using statistical methods like propensity score analysis and multilevel modeling. They focus on causal inference to evaluate different learning platforms.
What working here is like
- Computational
- No organisms (data only)
- Human participants
- Coding required
- Projects
- 7
- Papers
- 5
Current projects
Evaluating Virtual Learning Environments
Researching the effectiveness of different virtual learning platforms using causal inference methods.
Improving Early Literacy with AI: Fine‑Tuning GPT‑4o Mini for Phonics‑Based Curriculum
The team created a set of 1,290 short stories that match specific phonics lessons and used them to fine‑tune a smaller GPT‑4o model. The updated model now generates reading passages that follow the curriculum’s sound and word rules, and they evaluate it on story clarity, length, speed, and curriculum fit to help young readers learn to decode words.
From the Spring 2025 UF Undergraduate Research Symposium · abstract book, p. 347
Show 5 more projectsShow fewer
Identifying Features for Stealth Assessment of Oral Reading Fluency with AI
The lab uses AI to listen to kids reading aloud and picks out mistakes, creating measurements that predict how many words they read correctly per minute. The student built and tested speech features that help assess reading fluency.
From the Spring 2025 UF Undergraduate Research Symposium · abstract book, p. 152
Aligning AI-Generated Children's Stories with Reading Curricula
Researchers used machine‑learning and language‑processing methods to check how well AI‑written kids’ stories match the goals of a school reading program, creating a fast way to keep digital books on track with lessons.
From the Spring 2025 UF Undergraduate Research Symposium · abstract book, p. 37
Project READS - Reading Error Analysis and Detection System
The lab created an AI tool that listens to elementary students read aloud, spots pronunciation mistakes, and adjusts the story content to help them practice the sounds they find hard, aiming to boost reading fluency.
From the Spring 2025 UF Undergraduate Research Symposium · abstract book, p. 36
Project InSPIRE (Intelligent Support for Personalized Individual Reading Experiences)
The lab creates a digital version of a phonics tool called Elkonin boxes to help children with dyslexia, ADHD, or autism improve reading. The student builds the digital tool and evaluates its use in the app.
From the Spring 2025 UF Undergraduate Research Symposium · abstract book, p. 12
Project FUNLIT - Fun Literacy Integration for Thriving Readers: Gamification of Reading Practice to Increase Engagement and Motivation
The lab adds game elements like points and badges to a reading app to make kids want to read more. The student helps design the game features and tests whether they improve reading practice.
From the Spring 2025 UF Undergraduate Research Symposium · abstract book, p. 11
Summaries are written by AI in plain language from public symposium abstracts and may be imperfect.
Recent papers, explained
- Enhancing propensity score analysis with data missing not at random: Introducing dual-forest proximity imputation
2026 · Behavior Research Methods
Uses random forests to improve handling of missing data in treatment effect studies.
- Sensitivity Analyses for Omitted Confounders in Structural Equation Models with Tabu Search Optimization
2026 · Structural Equation Modeling A Multidisciplinary Journal
Uses search methods to check how unmeasured factors might affect study results.
Show 3 more papersShow fewer
- Understanding the heterogeneity of effects of teacher professional development: a systematic meta-review
2026 · Professional Development in Education
Reviews six studies to find what makes teacher training programs effective.
- Self-Regulation, Scaffolding, and the Illusion of Improvement: A Quasi-Experimental Evaluation of an LLM-Based Programming Help Tool
2026 · IEEE Access
Tests whether a coding help tool changes student learning behaviors.
- Machine learning for propensity score estimation: A systematic review and reporting guidelines.
2025 · Psychological Methods
Reviews how machine learning is used to estimate treatment effects.
Plain-language summaries are AI-written and reviewed by us; check the paper for details.
How to reach out
Walter Leite hasn't set up this page yet. You can also look them up in the UF directory (search “Walter Leite”).