Embedded Files
Introduction
Introduction
The ECOL 182L course is a research-based undergraduate experience (CURE) designed by Dr. Ryan Ruboyianes in which students conduct a novel scientific experiment and learn to report their findings. Students in ECOL 182L often struggle to apply statistical concepts, specifically to perform and analyze simple hypothesis tests.
Biology projects often require analyzing and interpreting data through statistical analysis. Biology majors vary in their previous knowledge of math, making it difficult to teach statistical analysis. Instructors spend a large portion of the session explaining basic statistical concepts. Students with stronger foundations lose interest in classes as they relearn material that they have already mastered. Meanwhile, students with weak math foundations often lag behind as they require a slower pace.
Tutoring is an effective way to homogenize student knowledge. Through one-on-one sessions, less-prepared students can quickly develop an understanding similar to that of their peers. Tutoring has several limitations. First, it is time-intensive. Usually, classrooms only have one instructor, and one-on-one discussions are hard to schedule. Second, because private tutoring is expensive, access for low-income students is limited, exacerbating disparities between peers.
AI tutoring has emerged as an alternative to private tutoring. AI tutors offer many benefits to student learning (Park University, 2025). First, AI tutors can personalize their content according to the student’s knowledge. Second, AI tutors offer immediate feedback to students. Timely feedback allows students to scaffold their learning, connect ideas before forgetting the material, and increases student motivation (Fisher et al., 2025).
TAR Question
TAR Question
Do AI tutors facilitate the design of hypothesis tests and the interpretation of statistical tests for biology undergraduate students?
Approach/Methods
Approach/Methods
We divided students into two groups: a control active-learning group and an experimental AI-tutor group. Both groups completed a four-question quiz on the design of a simple hypothesis test and the interpretation of a p-value. For the control group, I led a think-pair-share activity in which students worked in groups of 4. For the experimental group, I partnered with Jayaram Timsina, Enterprise IT Architect for Academic, Student, and Business Operations, to design an AI tutor. At the end of each activity, I led an in-class debrief and discussion.
After the activities, the students completed a survey reporting their feelings during both activities. After a week, the students completed a summative assessment containing two questions. To grade the assessment, I use a rubric that assigns a score from 0 to 2 for each question.
Results
Results
Students in both groups liked the two activities equally (Mann-Whitney test p-value=0.67 )
Students in the AI group performed similar to the active learning group than students in the active learning group (Chi-square test p-value=0.64 )
The active learning group performed better at interpreting p-values (p-value = 0.048) but not when sampling students who completed half ot the AI tutor activity.
Student feedback
Discussion / Lessons Learned
Discussion / Lessons Learned
AI tutors are as effective in teaching students statistical concepts as state-of-the-art active learning techniques. AI tutors might enhance student engagement by creating spaces where introverted students, who often feel intimidated by group participation, can actively engage with the course content. This makes AI tutoring an effective UDL technique.
Future work should explore how students interact with AI tutors and under what circumstances AI tutoring is more effective.
Acknowledgements
Acknowledgements
I want to thank Dr. Kristin Winet for her amazing guidance through this TAR project. I also want to thank Jayaram Timsina for his amazing help in designing and implementing the AI tutor. Finally, I would like to thank Dr. Joanna Masel and the Spring 2026 TAR cohort for their valuable suggestions.
References
References
Park University. (2025, February 14). AI in education: The rise of intelligent tutoring systems. Retrieved February 15, 2026, from https://www.park.edu/blog/ai-in-education-the-rise-of-intelligent-tutoring-systems/
Fisher, D. P., Brotto, G., Lim, I., & Southam, C. (2025). The Impact of Timely Formative Feedback on University Student Motivation. Assessment & Evaluation in Higher Education, 50(4), 622–631. https://doi.org/10.1080/02602938.2025.2449891
About the Author
About the Author
Ulises Hernández is a PhD candidate in the Department of Ecology and Evolutionary Biology. He studies the impact of genetic differences on life outcomes such as diseases. He is also interested in designing teaching practices that engage students with different learning styles. In his free time, he enjoys climbing, martial arts, and reading.
Page updated
Report abuse