Using data to understanding student performance in schools considering factors such as parental level of education, gender, lunch type and test preparation course.
As I explored the dataset using only Excel, patterns slowly began to reveal themselves: students who excelled not because they were “naturally gifted,” but because of support systems, preparation, nutrition, or parental education. Some succeeded quietly. Some struggled despite effort. And some soared once they completed the test preparation course. This report tells that story: It shows what truly influences the performance of Harvard University students—not just grades, but human factors like gender, race, family background, and access to learning opportunities. The goal of this analysis is to uncover evidence-based insights that any school—elite or not—can use to help students thrive.
- Feature Engineering: Created Total and Status columns.
- Excel Techniques Used: Pivot tables, slicers, SUM and IF statement and Pivot table Grouping.
- Business-Focused Insights: Identified strongest predictor of successful student performance and the importance of nutrition for students.
- No external tools used — the entire analysis was conducted using Excel to demonstrate strong fundamentals.
- Parental education is the strongest predictor of success.
- Test preparation improves scores dramatically (225 vs. 199).
- Females outperform males in reading and writing.
- Race Groups E & D excel; Group A needs intervention.
- Standard lunch students perform higher than free/reduced lunch students.
- Expand test prep access.
- Improve math teaching resources.
- Build equity support for lower parental-education groups.
- Strengthen nutrition programs.
- Establish enrichment for high performers.
Read full article on Medium: Beyond the Grades: An Excel-Driven Story of Harvard University Students’ Performance
Oyibo Dorcas
Data Analyst
Passionate about student performance, business analytics, and data storytelling.