Files
Abstract
Voluntary employee turnover, or attrition, can be a significant impediment to the effective functioning of an organization. With each employee replacement, time and money are spent to search for and train a new hire, so companies strive to identify areas that lead to turnover and effectively plan for its occurrence. Academics in organizational psychology and management studies have produced extensive literature on this issue, often taking the approach of identifying variables that range from employee-specific to industry-wide. However, results are not consistent at the company level, as each organization is placed in a unique context. As a result, companies benefit the most from conducting analysis using their historical data. The present study implemented this approach by determining the most significant drivers of employee turnover at a mid-size food and beverage organization (“the Company”). Given 6 years of data on employee attributes and turnover, this study used various interpretable machine learning models to point to key correlates of voluntary turnover in addition to generating predicted probabilities based on the current workforce. This analysis drove the main deliverable, presented to the Company in the form of a slide deck, which visualizes the relationships between significant variables and voluntary turnover. Additionally, turnover forecasts for 2024 were generated. The models used in this study showed strong predictive ability, with consistent variables emerging as significant across various model types. This analysis highlights areas of interest that are predictive of voluntary turnover while identifying further opportunities for inquiry.