We Stopped Relying on Teammates: LLM Use and Collaboration in a Student Hackathon

This README file describes the structure, content, and organization of the Hackathon Mixed Method(Qualitative and Quantitative) Analysis data repository. 1. Project Overview This repository contains data and materials from a study on how hackathon teams use large language models (LLMs) while working on team-based software development tasks. The study focuses on three teams (Team 1, Team 2, Team 3) and addresses two main questions: RQ1: How do LLM-supported collaborations impact team dynamics in hackathons? RQ2: How do participants perceive the role of LLMs across different hackathon activities? RQ3: How do participants perceive the usefulness and impact of generative AI on team dynamics and hackathon activities? Data include observation notes, pre/post survey responses, interview transcripts and qualitative coding files (codebook + axial codes). 2. File Descriptions Instruments Debrief_Interview_Script.pdf : Semi-structured interview guide used in post-hackathon debriefs. Observation_Guide.pdf : Protocol used by observers to structure field notes during the hackathon. Qualitative Data Observation notes (by team) team1_observation-notes_raw.pdf : Original handwritten/scan observation notes for Team 1. team1_observation-notes.docx : Typed/cleaned observation notes for Team 1. team2_observation-notes.docx : Typed observation notes for Team 2. team3_observation-notes.docx : Typed observation notes for Team 3. Interview transcripts (by team) team1_processed-interview-transcript.txt team2_processed-interview-transcript.txt Team3_processed-interview-transcript.txt (Each file contains a processed, pseudonymized transcript for that team’s debrief interview.) Survey Data pre-survey_responses.xlsx – Participants’ responses to the pre-hackathon survey. post-survey_responses.xlsx – Participants’ responses to the post-hackathon survey. Coding Files Codebook_v1.xlsx – Initial qualitative codebook (code names, definitions). Axial_Code_v1.xlsx – Higher-level categories built from the initial codebook, showing how codes were grouped into themes. 3. How to Use This Repository A few suggested entry points, depending on your goals: To understand the study design Read the Project Overview section above. Open Observation_Guide.pdf and Debrief_Interview_Script.pdf to see how data were collected. To explore the qualitative data Start with the team-level observation notes and interview transcripts (teamX_observation-notes, teamX_processed-interview-transcript.txt). Use Codebook_v1.xlsx to understand how segments were originally coded. Consult Axial_Code_v1.xlsx to see how those codes were aggregated into broader themes. To work with the survey data Load pre-survey_responses.xlsx and post-survey_responses.xlsx into your preferred stats environment. Conduct descriptive analyses. 4. Data Organization Files follow a team-first naming convention: team1, team2, team3 indicate which hackathon team the file belongs to. Data are grouped conceptually into: Instruments (guides/scripts) Qualitative data (observations + interviews) Survey data (pre/post) Analytic artifacts (codebook + axial codes) Participant identifiers in the qualitative data use pseudonymous IDs (P1, P2, etc.) rather than real names. 5. Ethics All identifying participant information has been removed or anonymized. This repository is for research and educational purposes only. Redistribution of raw or identifiable data is prohibited without authorization from the study leads. For questions or collaboration requests, contact the study coordinator or corresponding author.