Authors : P. S. Sudharshini; R. B. I. V. Rathnamalala; A. I. Udeshika; D. C. Dahanayaka

Volume/Issue : Volume 10 - 2025, Issue 11 - November

Google Scholar : https://tinyurl.com/yc8yw2mv

Scribd : https://tinyurl.com/4ns5yjj8

DOI : https://doi.org/10.38124/ijisrt/25nov1298

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : This exploratory data analysis (EDA) examines the factors influencing employee attrition using a Kaggle dataset comprising 1000 employee records with 26 demographic, job-related, compensation, and satisfaction variables. The target variable, Attrition (Yes/No), reveals a relatively low overall turnover rate of 18.9%, with a marked class imbalance (81.1% No vs. 18.9% Yes). Descriptive statistics and visualizations highlight a mid-career workforce (median age 41 years, average tenure approximately 15 years), male dominated gender composition, and finance as the largest department. Key findings indicate that lower monthly income, overtime work, and certain high pressure job roles (particularly Executive and Manager positions) are strongly associated with higher attrition rates. Employees who left the organization typically earned less, worked overtime more frequently, and had slightly shorter tenure, suggesting compensation dissatisfaction, burnout, and early career vulnerability as prominent drivers. Surprisingly, job satisfaction alone showed weak predictive power, with attrition occurring across all satisfaction levels, including high ratings, underscoring the multifaceted nature of turnover. While age and tenure provide contextual insights, compensation, overtime, and job role emerged as the strongest correlates of attrition. These results offer actionable insight for organizations to prioritize competitive salary structures, overtime reduction policies, and targeted retention programs for leadership roles, thereby reducing turnover costs and fostering a more stable, engaged workforce.

References :

1. Andrew T. Jebb., Scott Parrigon., & Sang Eun Woo. (2016). Exploratory Data Analysis as a Foundation of Inductive Research. Human Resource Management Review. 27(2). https://doi.org/10.1016/j.hrmr.2016.08.003
2. Ayesha Banu Mohd., Sharmila Reddy., & Rama M A. (2022). Exploratory Data Analysis (GEDA): A Case Study on Employee Attrition. 10.46243/jst. 2022.v7. i09.pp01-11
3. Fatbardha Maloku., Besnik Maloku. (2024). Analysing IBM HR Data: Employee Attrition and Performance Insights. Journal of Engineering and Applied Sciences Technology. 6(8): 1-10:1-10. https://doi.org/10.47363/JEAST/2024(6)268
4. Hardik I., Dharmendra Patel. (2024). Exploratory Data Analysis and Feature Selection for Predictive Modelling of Student Academic Performance Using a Proposed Dataset. International Journal of Engineering Trends and Technology. 72(11):131-143. https://doi.org/10.14445/22315381/IJETT-V72I11P116
5. Mohammed Salmanuddin., Rushikesh Kulkarani., Atharva Mohite., & Mahendra Patil. (2023). Exploratory Data Analysis. ICSTEM. https://doi.org/10.35629/5252-050413881392
6. Simon Gim., & Eun Tack Im. (2023). A Study on Predicting Employee Attrition Using Machine Learning. Studies in Computational Intelligence. https://doi.org/10.1007/978-3-031-19608-9_5
7. Kaggle Dataset: https://www.kaggle.com/datasets/ziya07/employee-attrition-prediction-dataset