影響我國15~29歲青年勞工工作轉換之因素探討

A study on influential factors associated with job turnovers in Taiwanese young labors

王韻婷(YUN-TING WANG)

指導教授 ： 黃怡婷

國立臺北大學/商學院/統計學系/碩士(2016年)

摘要

青年是國家的重要資產，其就業初期階段的經歷與選擇將牽動著個人及整體社會及企業的發展。然而，面對全球無疆界經貿市場與網絡形成的劇烈競爭、少子女與高齡化的人口結構等因素影響，導致國內產業結構調整，產業人力需求呈現兩極化趨勢，青年進入職場的就業環境較以往更形複雜，青年高失業率成為當今世界各國政府在發展人力資源政策最關注的議題。
　　本文利用勞動部 2014 年「15~29歲青年勞工就業狀況調查」共 3,963 筆資料，透過不同的統計模型探索影響青年勞工工作轉換的重要因子，並深入對教育程度在大專以上之青年受僱勞工建模分析。本研究之主要模型建置分為二種：「工作流動模型」的反應變項為「轉職次數」，惟資料中「零值」次數相當多，因此除採用卜瓦松迴歸方法，亦使用零膨脹卜瓦松迴歸方法配適；「轉職意願模型」的反應變項為「目前是否有轉職打算」，採用二元邏輯斯迴歸方法配適。
　　實證結果顯示，在工作流動模型中，青年勞工工作轉換次數存在零膨脹現象，而大專畢業之青年勞工工作轉換次數則無，共同有影響的因素包含「獎金待遇、學以致用程度、薪資差異、初職工作型態、初次尋職遭遇困難情形、參與教育訓練、持有證照情形、全國就業 e 網或台灣就業通、年齡」，惟大專畢業之青年工作者較整體青年工作者不受「工作職類、員工規模、尋職方式、初職尋職時間、教育程度、每月薪資主要運用」的影響，而另以「主修科系」對轉職情形有所影響。
　　在轉職意願模型中，影響整體青年勞工及大專青年勞工之工作轉換的共同因子，包含現職工作型態、現職契約性質、每週工時、加班情形、獎金待遇、學以致用程度、月薪、現職工作年資、參與教育訓練、持有證照情形、就業博覽會現場徵才活動、總工作年資，惟大專畢業之青年工作者較整體青年工作者不受「工作職類、教育程度、轉職次數」的影響，而另以「年齡」對其轉職意願有所影響。值得一提的是，不論學歷程度為何，「加班情形」為所有顯著因素中最重要的影響因子。本論文的結果可提供規劃促進青年就業政策及學術單位研究分析應用。

關鍵字

邏輯斯迴歸 ； 零膨脹卜瓦松迴歸 ； 卜瓦松迴歸 ； 轉職次數 ； 轉職意願

並列摘要

Young labors are important assets for a nation. Their job experience and career choice initially influence development simultaneously in the aspects of individual, society and business. Nevertheless, owing to globalization in business, and unbalanced population structures, industries have to adjust their manufactures and productions, which result in laboring adjustment. It then becomes much complex to find the first job for young people. The high unemployment in youths becomes an important issue when making the policies for sufficiently used manpower.
This study used data of the youth employment labor survey collected by Ministry of Labor in 2014. There were 3,963 eligible respondents. Two different models were discussed to explore the important factors affecting the young labors for changing their jobs. To have a specific understanding about college graduates, two more models are built. The first model discusses the job mobility, which is measured by the number of job-changes. Since there is a high frequency in zero job change, two statistical models including Poisson regression model and Zero-inflated Poisson regression model were used. The second model was the intention for turnover, which is measured by a binary variable, “willingness to transfer job now”. The logistic regression model was used to assess the influential factors.
The empirical results show that for the job mobility, Zero-inflated Poisson is performed well, whereas the regular Poisson model has the best performance for the college graduated. The common impact factors in both models contain bonus, level of education applied, wage gap, nature of first job, the obstacle for finding the first job, educational training, the number of licenses, ejob/taiwanjobs-known, age. The significant factors only for the overall model include occupation, the number of staffs, the way of finding a job, the number of times when seeking the first job, level of education, major usage of salary. In addition, the major of college is significant for the model for college graduates.
For the intention for turnover, the common impact factors contain nature of job, contract of job, weekly working hours, overtime, bonus, level of education applied, wage, years of job, educational training, the number of licenses, job fair event and total work years. The difference between overall and college graduated young labors are occupation, level of education and the number of turnovers. Age is only significant in the college graduates model. In particular, overtime is the most influential factor among all significant factors for both models. The result of this thesis can be used to make policies for improving young employment and provide insights for academic research.

並列關鍵字

Number of Job-changes ； Logistic Regression ； Zero-inflated Poisson Regression ； Poisson Regression ； Intention for Turnover

參考文獻

蔡怡萍 (2013), 台灣青年就業形式與自願性選擇探究, 國立臺灣大學國家發展研究所碩士論文。

陳又綾 (2008), 影響年輕族群薪資與離職傾向之因素探討, 國立中央大學人力資源管理研究所碩士論文。

林詩敏 (2006), 中年女性職場轉換之經驗與學習研究－以兒童補教業為例, 國立臺灣師範大學社會教育學系碩士論文。

Agresti, A. (2006). “An Introduction to Categorical Data Analysis”, Second Edition, John Wiley & Sons, New York.

Agresti, A. (2012). “Categorical Data Analysis”, Third Edition, John Wiley & Sons, New York.