This project studies how classroom size may affect kindergarten students' math performance.
We use data from the Tennessee STAR project, which is a well known education study about class size.
Our goal is to compare math scores between three classroom types:
- Small classes
- Regular classes
- Regular classes with a teacher aide
The analysis was completed as part of the STA 207 final project at UC Davis.
Our main question is:
Does class size influence kindergarten students' math performance?
We are especially interested in whether students in smaller classes perform better than students in regular classrooms.
The dataset comes from the Tennessee STAR (Student/Teacher Achievement Ratio) Project.
This experiment started in 1985 and randomly assigned students to different classroom types.
Because the assignment was randomized, it helps us study the relationship between class size and student performance.
For this project, we focused on:
- Kindergarten students
- Kindergarten math scores
- Classroom type
Our analysis includes several steps.
We first explored the data to understand the distribution of math scores.
We used summary statistics and visualizations to compare classroom types.
We compared math scores across the three classroom types to see whether class size is related to student performance.
We also checked whether the results remain similar under different model settings.
Small classes show better performance
Kindergarten students in small classes tend to have higher math scores compared with students in regular classrooms.
Teacher aides do not show a strong effect
We did not find a clear difference between regular classes and regular classes with a teacher aide.
Limitations
Our analysis only focuses on:
- Kindergarten students
- Schools in Tennessee
- Data collected beginning in 1985
Because of this, the results should not be generalized to all schools or all grade levels.
Future research could extend this work by:
- Studying higher grade levels
- Looking at long term student outcomes
- Using newer education datasets
- Including schools from more regions
project/
│
├── code/
│ R scripts used for data analysis
│
├── report/
│ Final HTML report
│
├── figures/
│ Visualizations used in the report
│
└── README.md
- Chi-Chun (Emily) Chen
- Elaine Guo
- Skylar Calandra
- William Baggott
University of California, Davis
STA 207 – Statistical Methods for Research II