Exploring Predictors of Turnover for New Employees Using Decision Tree Analysis: A Comparative Study of Gender Differences

The purpose of this study is to comprehensively explore the factors that influence new employees' turnover intention and compare whether there are differences in the influence of each factor on turnover intention based on gender. Based on the results, suggestions were made to reduce the turnover rate of new employees. In this study, decision tree analysis was used as a machine learning technique, and turnover intention was set as a predictor variable, and the explanatory variables were composed of 26 variables belonging to personal factors, organizational factors, and job-related factors based on previous studies. When examining the top variables that affect turnover intentions of new employees, we found that variables such as 'job satisfaction', 'job-education match', 'working status', and 'age' are important. We also found that there are differences in the factors that affect turnover intentions of new employees between genders. For male new hires, 'job satisfaction' was found to be the most influential, followed by 'job-education level match', 'working status', 'age', 'HR system satisfaction', 'job classification', and 'major field of study'. In the case of female new hires, 'match between job and skill level' was found to be the most influential, while variables such as 'job satisfaction', 'working status', 'labor union', 'business type', and 'number of employees' were found to be important in considering a job change. The results suggest that a strategic approach to gender-based staffing is necessary, and at the same time, the turnover of new employees should be addressed in consideration of the fact that the MZ generation has a different concept of workplace than the previous generation.