More Than Merit: The Factors Behind Standardized Test Scores

​Standardized tests like the SAT are a key part of college admissions in the US and are widely taken by high school students across the nation. Many students work hard to try and achieve high test scores to be competitive in college admissions. Merit may not be the only factor influencing test scores, though. Factors outside of a student’s control may also play a role in their test scores. For example, schools with insufficient resources may not always have access to standardized testing tutoring services or a personalized training for students. This project attempts to determine correlations between school indicators and average SAT results. The high schools analyzed in this project are located in Illinois. The main factors that will be analyzed in this project are: school population, percentage of attending students considered low-income, average annual state funding per student, student-to-teacher ratio, graduation rate, and region. As of 2021, all high schools in Illinois require juniors to take the SAT, creating a data point which can be easily accessed, as well as an objective variable to analyze. The purpose of this project aims to see if there is a correlation between various school factors and standardized testing scores.

 

Our group gathered data on SAT score from Niche.com and data on all other factors from each school’s Illinois Report Card page. We tested correlations between SAT score and school region, low-income rate, spending per student, and student-to-teacher ratio. We ran paired t-tests and correlation calculations on the school data gathered in Microsoft Excel. Our group operated under the assumption that there was no significant difference between SAT scores and the measured variables. In other words, the relationship between SAT scores and the measured variables is due to chance. We used p-values from the paired t-tests to determine whether there were significant differences between SAT scores and the measured variable. If the p-value is under 0.05, then there is a significant difference. We used R-values to determine the strength of the correlations between SAT scores and the analyzed factors. Our data visualizations were made with Plotly and Javascript.
 

​