Application of the Two-stage Cluster Analysis on Employee Voluntary Turnover Intention

• Main Authors: Hsiao-Ju Chien, 簡筱茹
• Other Authors: Pei-Chann Chang
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/08666912236013136701

碩士 === 元智大學 === 工業工程與管理學系 === 95 === Utilizing human resources adequately and converting organizations into learning or-ganizations are the sources of advantage in long-term competitions of industry. The mo-tivation of this study is to explore brain drain and falls of techniques in the future of Taiwan, so as to control and make preparations. Most researchers review the behavior of departee, and forecasts of leaving tendency are usually overlooked; so, resulting in not avoiding brain drain efficiently and immediately. Consequently, cluster analysis of data mining is utilized to resolve the problem; the methodology includes the combination of Self-Organizing Map and Cluster Analysis to examine the individual characteristic of turnover intention. The subjects of questionnaires investigation are persons who are em-ployed in famous companies and office holders across the north, central, south regions of Taiwan, and persons who are in the boom season of turnover in the Chinese New Year. The number of total questionnaires are 605, and 421 copies are available. The findings present the marked characteristics of turnover intention; such as, not identifica-tion with supervisory loyalty, guidance and management. Therefore, it does not present the common cognition about unsatisfying with salary and welfare. The accuracy of Cross Validation is 92.7％, the ability of sample categorization and differentiation are up to 100％. According to the findings, the forecasting model of turnover intention is established; it can be a reference of management of employee turnover, and makes a contribution to prevent brain drain, stimulating organization learning opportunities, promoting competing abilities of industries.