T.I.P.S. Cohort Insights
Project Details
Overview & Team Collaboration
The T.I.P.S. Project was a team-based cohort analysis. My team processed a historical dataset of student advice spanning ten cohorts (2016–2026) of the NC State Master of Science in Analytics (MSA) program. We combined qualitative text analytics with spreadsheet modeling, transforming unstructured advice into a clean, visual roadmap for future cohorts.
Methodology
Our analytics approach focused on two main components:
- Excel Cohort Tracker: Compiled historical class metrics, graduation numbers, placement rates, and salaries into a structured Excel timeline to trace program outcomes over a ten-year span.
- Natural Language Processing (NLP): Built a Python pipeline utilizing sentence embeddings (
all-MiniLM-L6-v2) to map each piece of written student advice into a semantic vector space. This allowed us to measure sentiment trends, determine the semantic center (most representative advice), and identify unique outliers.
Key Findings & Takeaways
The Central Theme of "Time"
Through semantic center modeling of the entire dataset, we found that the concept of time was the most critical recommendation across all ten cohorts. MSA predecessors consistently highlighted the importance of three key areas:
- Time Management: Balancing the fast-paced program workload.
- Peer Connection: Dedicating quality time to build relationships with classmates.
- Faculty Engagement: Actively taking time to connect with faculty and staff.
The Shift in Programming Tech
Our timeline analysis revealed a clear, industry-driven evolution in the software recommendations made by students:
- SAS & R: Fell from high prominence in 2016 to near-zero mentions by 2026.
- Python: Remained a steady, highly recommended staple throughout the decade.
- SQL: Saw a huge rise, emerging as the primary technical skill recommended by 2026.
- AI Toolsets: Appeared exclusively in the 2026 tips, with roughly 20% of the class advising students on how to responsibly integrate generative AI tools into their workflows.
Surprise Outliers: Student Food Culture
By mapping tips that were furthest from the semantic center, we discovered the most unique, outlier tips. Interestingly, they were almost entirely centered around Raleigh's local food culture, ranging from cafe ordering hacks to free food strategies on campus.
The End Result: The T.I.P.S. Framework
We synthesized our findings into a clean, four-pillar framework designed to guide incoming analytics students:
- T - Time: Master time management, and prioritize relationships with peers and faculty.
- I - Insights: Use cohort data to understand placement and curriculum trends.
- P - Programming: Focus heavily on Python, SQL, and emerging AI tools.
- S - Surprises: Embrace the student community and local culture.