The main objective of this project is to conduct a comprehensive analysis of social media user behavior, with a particular focus on understanding patterns of daily activity and the dominant emotions exhibited by users. By examining key metrics such as daily usage time, number of posts, likes received, comments received, and messages sent, this study aims to uncover correlations and trends that reveal how users interact with social media platforms.
Additionally, the analysis explores the relationship between user activity levels and emotional well-being, providing insights into how engagement on social media may influence or reflect a user’s emotional state. Through the application of statistical techniques, data visualization, and exploratory data analysis, this project seeks to provide a data-driven understanding of social media behaviors and their potential psychological implications.
⏱ Daily Usage Time: Duration of social media usage per day
📝 Posts per Day: Number of posts made per day
❤️ Likes Received per Day: Number of likes received per day
💬 Comments Received per Day: Number of comments received per day
📩 Messages Sent per Day: Number of messages sent per day
😄 Dominant Emotion: User’s dominant emotion
Social Media Usage and Emotional Well-Being (Kaggle)
-
Programming Language: Python
-
Libraries & Tools:
-
pandas, numpy → data manipulation
-
matplotlib, seaborn → data visualization
-
scipy, statsmodels, researchpy → statistical analysis
-
scikit-learn → modeling and evaluation
1️⃣ Daily Activity Correlation
First correlation heatmap shows that Daily_Usage_Time, Posts_Per_Day, Likes_Received_Per_Day, Comments_Received_Per_Day, and Messages_Sent_Per_Day are highly correlated (0.86 – 0.94). → The longer users spend on social media, the more active their interactions.
Age shows low correlation (0.03 – 0.12) with all variables, indicating that age has no significant impact on social media activity.
2️⃣ Distribution & Variable Relationships
Pairplot / Scatterplot Matrix reveals a linear pattern between Daily_Usage_Time and Posts_Per_Day, Likes_Received_Per_Day, Comments_Received_Per_Day, and Messages_Sent_Per_Day.
The distribution of Age is relatively uniform, with no extreme outliers and no strong relationship with daily activity.
Some variables, such as Daily_Usage_Time, are skewed, with more users having moderate usage than extreme high usage.
3️⃣ Advanced Correlation
Second correlation heatmap confirms a high correlation between Likes_Received_Per_Day and Comments_Received_Per_Day (0.94).
High correlation is also observed between Daily_Usage_Time and interactions (likes & comments).
Correlation with Age remains low, reinforcing the conclusion that age does not significantly influence social media activity intensity.
4️⃣ Additional Insights
Most frequently used platforms: Instagram and Snapchat (average > 120 minutes per day).
Users with more posts typically receive more likes and comments.
Most users exhibit Neutral or Happiness emotions, with a smaller portion experiencing Anxiety or Sadness.
Higher activity tends to correlate with positive emotions, although some highly active users also show high levels of anxiety.
5️⃣ Visualizations
Correlation heatmap of activity and emotions
Histogram of usage time per platform
Scatter plot of posts vs likes
Conclusion
Daily social media usage duration is a strong indicator of user activity.
Interaction activities (posts, likes, comments, messages) are highly correlated.
Age has no significant effect on the intensity of social media activity.
Findings are consistent with literature and the Kaggle dataset used.
-
Susan Jong (231401014)
-
Clarissa Halim (231401020)
-
Michael Purba (231401053)
-
Steven Winarta Girsang (231401059)
-
erarine (231401095)