This project explores student performance across multiple dimensions, including gender, ethnicity, parental education, and test preparation. By analyzing real-world student exam scores, the project seeks to understand key factors affecting academic outcomes.
Dataset Source: Students Performance in Exams
Despite increasing educational initiatives, student performance often varies due to demographic and socio-economic factors. The goal is to:
- Analyze student score trends in math, reading, and writing.
- Identify performance gaps across gender, parental education, and ethnicity.
- Provide actionable insights for institutional effectiveness.
- Project Overview
- Data Understanding
- Data Cleaning
- Exploratory Data Analysis (EDA)
- Insights Derived
- Suggestions
- Challenges Faced
- Future Scope
- Final Outcome
- SQL Analysis File
- Project Credits
This analysis dives into students' exam performance and highlights influential factors like:
- Gender differences in scores
- Group-wise academic achievements
- Influence of parental education and test preparation
- Dataset has 1000 entries with 8 categorical and 3 numerical fields.
- No missing values, but categorical columns required renaming and restructuring.
- Additional derived fields:
avg_scoreandgrades.
- Renamed columns for clarity (e.g.,
math score→math_score) - Added:
avg_score: average of math, reading, and writinggrades: A+ to D, based on average score
-
Gender Differences
- Girls scored higher in reading & writing.
- Boys slightly outperformed in math.
-
Test Preparation
- Students completing test prep scored significantly higher.
- Strong correlation with better grades.
-
Parental Education
- Higher parental education led to better student scores.
- Students of graduate-educated parents performed best.
-
Ethnicity Groups
- Group E performed best, Group A lowest.
- Socio-economic influence possible.
-
Grade Distribution
- Most students scored in grade B or C.
- Very few in A+.
-
Subject Relationships
- Strong correlation between reading & writing.
-
Test Prep Works
- Students in test prep programs score higher.
- Expand access or make it mandatory.
-
Gender-Based Gaps
- Girls excel in reading/writing; boys slightly lead in math.
- Offer subject-specific support by gender.
-
Parental Education Impact
- Higher parental education = better student scores.
- Provide extra help to students with less-educated parents.
-
Group Performance Gaps
- Some ethnic groups consistently underperform.
- Launch group-specific academic support.
-
Grades Cluster at Mid-Level
- Most students are in B/C range.
- Refine curriculum to lift more students to A levels.
- Balancing label density in pie and bar charts.
- Ensuring the EDA output was visually understandable without being overwhelming.
- Creating a grading logic that fairly represented performance.
- Add predictive modeling to estimate student performance based on demographic factors.
- Apply clustering to segment students for targeted support.
- Integrate more diverse datasets from multiple schools or states for comparative analysis.
This analysis revealed:
- Gender and parental education have a measurable impact on performance.
- Institutional programs like test prep significantly improve outcomes.
- Data visualization helped stakeholders better understand performance trends and make decisions backed by evidence.
You can find the SQL queries and logic used for backend analysis in the following file: