랜덤 포레스트를 활용한 대졸 신입사원 조기이직 예측 결정요인 탐색

초록

This study aims to provide basic grounds for reducing the early turn over rate of university graduate newcomers. For this purpose, this study developed a random forest model for the prediction of university graduate newcomers’ early turn over using data from the Graduates Occupational Mobility Survey by the Korea Employment Information Service(2016GOMS). In the prediction model, 30 variables consisting of individual characteristics, job-related, organizational structural and psychosocial factors were applied. The top 10 factors predicting early turn over in the first jobs were initial salary, the severance payment, workplace satisfaction, job security satisfaction, the presence of labor union, work environment satisfaction from organizational structural factors and job satisfaction, job-aptitude fit from job-related factors and relationship satisfaction, possibilities for the personal growth of psychosocial factors. The result of this study demonstrates that good quality jobs with stability should be secured to support a smooth transition from school to work. It is also suggested that career education enhancing self-perception and job reality be proceeded at the aspect of education and social supports helping newcomers settle into working environment be offered at the organizational level.

키워드