The purpose of this study is to determine what variables that teachers report to be most reliable predictors of the potential dropout. Subsequently, it provides knowledge for the formulation of appropriate intervention strategies. This is a survey study. 23 teachers not knowing the aims of survey were invited to report their perceptions of 550 students from two high schools in a formal survey. Then, another teacher survey was then applied to identify whether students had become school dropouts or not. Chi square, Phi correlation, multinomial logistic regression, binary logistic regression and log-linear modeling were used to analyze the results. The findings from this study suggested five variables are the most reliable predictors reported by teachers to identify the potential dropout (i.e. poor academic performance, part-time job, absenteeism, suspension and family structure risk). Furthermore, five combinations are most associated with school dropout (i.e. absenteeism with discipline problem, absenteeism with tardiness, work with tardiness, discipline problem with tardiness, absenteeism with suspension). Besides the combinations, six other predictors are statistically reliable in predicting student's study status (i.e. primary education completion in Macao,absenteeism, school transfer, suspension, poor academic performance and part-time job). Apart from poor teacher relationship and poor peer relationship, all characteristic variables are statistically significant and associated with school dropout; and there is a significant difference between dropouts and persisters. Keywords: Dropout, Academic performance, Learning motivation

本研究以老師的觀點，探究澳門高危離校生之顯著心理與行為特徵，使為前線工作者提供適切的介入方案的相關條件。 本研究為調查研究，先利用文獻探討及前測來建立老師的紙筆問卷;正式施測則向 23位來自兩所中學之班導師進行調查，讓他們在未知本研究目的之情況下，感知報告其班上550名學生之顯著特徵。及後再填答學生之就學情況，以避免負面議題影響感知之客觀性。而結果分析利用卡方及相關考驗、多元類別邏輯迴歸分析、二元邏輯迴歸分析及階層對數線性模式分析。 結果顯示顯著能檢核中輟高危學生之預測變項分別有五:學科成績低、兼職、無故缺席、被停學、家庭危機;及五種與輟學相關之交互作用:缺席與紀律、缺席與遲到、兼職與遲到、紀律與遲到、缺席與停學;此外，於澳門完成整個小學學歷、缺席、轉學、被停學、學科成績低及兼職等六項能顯著預測學生之學習狀況;最後，本研究十七項心理與行為特徵中，除師生關係及同儕關係外，其他十五項特徵與輟學存在顯著相關，且輟學生與一般生在其他十五項特徵上有顯著差異。 關鍵詞:輟學生、學科成績、學習動機