Student Habits vs Academic Performance

Assignment 3 - Storytelling with Open Data

Many people believe things like “more study always means better marks” or “social media always reduces performance.”
This analysis uses a dataset of 1,000 simulated students from Kaggle to explore how daily habits relate to exam scores.
The dataset includes attributes such as study hours, sleep hours, social media use, Netflix time, attendance, exercise, and more.
The target variable is exam score (0–100).
The goal is to uncover clear, data-driven insights about which habits show the strongest relationship with exam performance and what daily patterns look like for lower vs higher scoring students.

The two visualizations below show the same information in different formats.
They reveal how strongly each habit is correlated with exam score.
The goal is to identify which habits have the highest positive impact on the exam performance and which habits may have a negative or minimal effect.

As seen in the previous slide, study hours per day have the strongest positive correlation with exam score.
The visualization below shows the same pattern clearly, the trend line rises steadily. The more a student studies, the higher their score tends to be.
This becomes especially obvious for students studying 6 hours or more per day, where most scores are above 75.
The goal here is to test the common belief that “studying more leads to better results”, and the data strongly supports it.

To understand how similar students are in their daily routines, we look at three common habits: study, sleep, and screen time.
Study hours and sleep hours follow a normal, balanced pattern, most students fall around healthy midpoints.
Social media hours are much more varied, with many students spending a lot more time on screens than others.
The goal is to see whether students are generally balanced, overworked, or distracted. Overall, students appear fairly balanced, but social media use shows the biggest imbalance.

We often believe that better sleep leads to better performance, and that being well-rested helps us stay focused during exams.
Here, we check how accurate this belief is and how strongly sleep hours actually affect exam scores.
Turned out more sleep tends to slightly improve exam performance, but the effect is much weaker and more scattered compared to the clear impact of study hours.
The goal is to see whether there is an optimal sleep range and to understand how different sleep durations relate to exam performance.

Attending classes regularly might show commitment, but does higher attendance always mean better marks?
I grouped students into low, medium, and high attendance and compared their exam scores.
The median score is highest for high attendance, then medium, and lowest for low attendance, but the three groups still overlap a lot.
This suggests that attendance has a positive effect, but it is not as strong or decisive as study hours per day.
The goal is to see how important attendance really is and whether it plays a supporting rather than a vital role in performance.

There is a common belief that more screen time leads to worse exam performance. Here we check how true this is, and how strong the effect might be.
For social media hours, the trend is slightly downward. Students who spend more time on social media tend to score a bit lower, but the data is very scattered. Many students with the same social media use have very different scores.
This means social media has a small negative effect, but it is not a strong or decisive factor on its own.
For Netflix hours, the negative pattern is clearer. As Netflix time increases, exam scores generally decrease, and heavy use (3–5+ hours) is more often linked to lower performance.
Netflix shows a moderate negative association, stronger than social media but still weaker than study hours.
The goal is to understand whether these screen-time habits have a meaningful correlation with performance, and how strong these effects are.

Here, we group students into high, medium, and low performers and compare all key habits across these groups.
High performers consistently show more study hours, slightly more sleep, slightly better attendance, and lower screen time.
Habits (sleep, attendance, social media, Netflix) show differences, but their effects are smaller and more overlapping.
Study hours remain the most impactful habit, but this combined view shows that performance is influenced by multiple small habits working together.

Study Hours: Strongest positive relationship with exam score; more study generally means better performance.
Sleep Hours: Moderate positive effect; consistent 6–8 hours supports better scores but the impact is modest.
Social Media Hours: Small negative relationship; more time on social media slightly lowers average scores but not strongly.
Attendance: Moderate positive effect; higher attendance is associated with higher median exam scores.
Netflix Hours: Small negative relationship; heavier Netflix use is more often linked with lower performance.

Dataset:
Nath, J. (2023). Student habits vs academic performance [Dataset]. Kaggle. https://www.kaggle.com/datasets/jayaantanaath/student-habits-vs-academic-performance