Analytics for Good Hackathon 2025 (WINNER!)

Fourth Annual Analytics for Good Hackathon

🎓 Boosting Student Engagement with Product Analytics

A data-driven case study analyzing the effectiveness of TABot mobile app, a smart study assistant used in a university setting. We investigated how engagement impacts academic performance and developed targeted recommendations using segmentation, visual insights, and behavior modeling.

---

🚀 Problem Statement

TABot helps students stay organized through reminders, study plans, and a chatbot interface. But does it truly impact learning outcomes?

---

🔍 What We Did

• Segmented students into:
  • Users vs. Non-Users
  • Active vs. Inactive Users
  • High vs. Low Performers
• Assessed differences in:
  • Assignment scores
  • Test scores
  • GPA
  • Canvas pageviews & submissions

---

📊 Key Insights

• TABot users outperform non-users on most academic metrics
• High activity users had significantly better outcomes than inactive users
• Students tend to lose interest mid-semester → retention challenge
• Identified four user archetypes based on performance and activity

---

💡 Recommendations

1. Data-Driven Personalization

• Custom study plans for low performers
• Personalized nudges and emails for inactive students

2. Gamification

• Add Duolingo-style mechanics: streaks, avatars, quick quizzes
• Use a reactive mascot ("Gopher Moods") and leaderboard-based challenges

3. UI/UX Improvements

• Improve chatbot responsiveness
• Enhance dashboard with progress tracking and personalized tips

---

🛠️ Tools & Skills

• Python (Pandas, Seaborn for EDA)
• Segmentation analysis
• Statistical significance testing
• Data storytelling and slide design

---

👥 Team TensionFlow

A cross-functional team of 6 MSBA students applying real-world analytics to improve student outcomes.

---

📂 Presentation: TensionFlow-PPT.pdf