Abstract

This research explores the determinants of employee burnout through Exploratory Data Analysis (EDA) and correlation analysis using the Synthetic HR Burnout Dataset from Kaggle. The study aims to quantify how psychological, workload, and demographic variables collectively influence burnout risk in simulated agile-team environments. A total of 2,000 employee records comprising ten features—including Age, Gender, JobRole, Experience, WorkHoursPerWeek, RemoteRatio, SatisfactionLevel, StressLevel, and Burnout—were analyzed using Python 3.11 with pandas, matplotlib, seaborn, and scipy. The workflow involved data cleaning, descriptive statistics, univariate visualization, and pairwise correlation computation (Pearson and point-biserial coefficients). Results show that burnout prevalence is 6.45 percent. The highest positive correlation with burnout occurs for StressLevel (r = 0.321, p < 0.001), followed by WorkHoursPerWeek (r = 0.226, p < 0.001), while SatisfactionLevel displays a moderate negative association (r = –0.233, p < 0.001). Demographic factors such as age, experience, and gender present statistically insignificant effects, suggesting that burnout is fundamentally behavioral and situational. Visualization outputs—including histograms, boxplots, and correlation heatmaps—reinforce these findings, revealing clusters of high stress and low satisfaction among burnout cases. From an Agile-management perspective, the study highlights that sustainable team performance depends on balancing workload and psychological well-being. The proposed analytical pipeline is fully reproducible on standard hardware, demonstrating that meaningful HR analytics can be conducted without large computational infrastructure. This approach provides a practical foundation for data-driven monitoring of team health, enabling organizations to translate the Agile value of “sustainable pace” into measurable, actionable metrics. Future research may extend this framework with predictive or time-series models using real HR or agile sprint data to explore burnout dynamics over time.

Keywords: Burnout Prediction, Exploratory Data Analysis, Correlation Analysis, Agile Management, Employee Well-Being

How to Cite:

Aziz, A. & Seno, A. S., (2026) “Investigating Employee Burnout Determinants through Exploratory Data and Correlation Analysis”, Agile Management 1(1), 17-30. doi: https://doi.org/10.63913/am.v1i1.124