本研究目的在比較台北市與台東縣兩地間國小教師工作滿意的差異並以社會網絡、工作特性為中介變項探究其因果機制，以期為影響兩地教師工作滿意的因素找到更有力的解釋及更確切的了解。本研究以分層叢集隨機取樣，抽取台北市與台東縣國小教師共906人為樣本，以自編問卷為工具，以因素分析及多元迴歸進行路徑分析。結果發現：女性、級任、兼行政職務、學歷較低者及收入較高的教師，工作滿意較高。而地區、族群、婚姻狀況對工作滿意的影響不顯著。整體而言，背景因素對工作滿意的影響相當小（R Square＝.04），在加入工作特性中的外在報酬、工作無意義、工作繁重及人際關係（亦屬社會網絡變項）等四個中介變項後，R Square提高為.31，顯示這四個中介變項具有相當大的解釋力。其中外在報酬及人際關係會提高工作滿意，而工作無意義及工作繁重會降低工作滿意。台東縣與台北市國小教師工作滿意沒有顯著差異，原因在於地區對中介變項的影響正負值相互抵銷所致。 本研究並對政府政策、學校與教師實務及未來研究提出建議。

The main purpose of this study is to develop a causal model to describe elementary school teachers’ job satisfaction and how it is affected by such mediating factors as social networks and job characteristics. By comparing the job satisfaction and its influencing factors of teachers who live and work in Taipei Municipality and those of teachers who live and work in Taitung County, it is expected that more powerful explanations and more exact understandings could be developed. Research participants were selected using the stratified cluster random sampling. Nine hundred and six elementary school teachers from Taipei Municipality and Taitung County were effective samples. The instrument for the study was the Questionnaire on Elementary School Teacher’s Work and Life Situations, developed by the researcher. In analyzing the research data, factor analysis and multiple regression were used. The study followed the procedure of path analysis to examine the causal mechanism behind the way background variables affect independent variables through the mediation of intervening variables. Major findings of the study can be summarized below: In terms of job satisfaction, the data indicats that higher levels of satisfaction are generally found in female teachers, homeroom teachers, teachers who assume directorship in the schools, teachers with lower educational attainment, and teachers with higher income. On the other hand, work locality, ethnic origin, and marital status did not seem to affect teachers’ job satisfaction significantly. However, the overall amount of influence these background variables had on teachers’ job satisfaction was minimal, since the R Square was only .04. Most of these differences became insignificant or decreased after four intervening variables related to job characteristics were added to the formulation. This indicated that the influence the aforementioned background variables had on job satisfaction was exerted through the mediation of these intervening variables. Among hem, interpersonal relationship ( also as a variable of social networks ) and external reward were positively linked to job satisfaction, and work meaninglessness and heaviness of work load were negatively linked to it. At the same time, the R Square took a significant increase to become .31, which indicated that they had significant explaining power. The influence of work locality was not significant, that was canceled out by some positive and negative effect of intervening variables. Suggestions for government policy, school and teacher practices are also included in the study for further study.