本研究探討：（1）見於大學生的，思考，人際和生活習慣與心理健康間的顯著關係是否也可見於高中生？（2）在高中生階段，讀書習慣是否也和心理健康有關？（3）那一種習慣和心理的關係較密切？（4）使用複項習慣回歸相關法預測心理健康時，其預測力有多高，且高於單項習慣的預測力多少？本研究使用高中一、二年級生（男674名，女603名，共1277名），大學生（共1602，男824名，女776名）為研究對象。所使用的工具是：（1）柯氏性格量表（1985），（2）短式的思考、人際、生活習慣量表，和讀書習慣量表。各量尺的信度，除了生活習慣量表之外，均良好。所得的結果是：（1）在高中生階段，思考、人際、生活等習慣仍然分別和心理健康有密切關係；（2）在高中生階段，讀書習慣和心理健康的關係也很密切；（3）思考與生活習慣比人際關係和心理健康形成更爲密切的關係；預測心理健康時，良好習慣比不良好習慣更有效，而預測消極面的心理健康（亦即症狀的多寡）時，不良習慣比良好習慣更有效;（4）多項的比單項的回歸相關法更能正確地預測心理健康；（以單項習慣最多只能解釋百分之三十七的心理健康分數變異量，但以複項習慣則可解釋的變異量會高達百分之四十與五十之間。

The study investigates the following questions: (1)If the intimate relations between habits and mental health found among University students also exist among senior high students? (2) If the Study Habit is also significantly related with mental health? (3) What kinds of habits are more closely related to mental health? (4)To what degree is multiple regression analysis method superior to ordinary single item correlational method in predicting mental health status from habits? How much portion of mental health variance could be accounted for by each of these two correlational methods, respectively? The study adopted two groups of subjects; namely, (1) 1277 healthy senior high students (674males, 603 females) ; and (2) 1602 healthy university students(826 males, 776 females). The tools used are the Ko's Mental Health Questionaire (1985), and four habits scales, including the Thinking Habit, the Interpersonal Relation Habit, the Daily Life Habit, and the Study Habits Scales. The findings are as follows: (1) the intimate relations between habits and mental health obtained in the previous study with university students also clearly exist among senior high students,(2) Habit of Study is also significantly related to mental health,(3) Habits of Thinking, Daily Life and Interpersonal Relations are important in predicting mental health, according to the order of importance here listed,(4)Multiple regression analysis is superior to simple correlational analysis in the prediction of mental health, indicating that mental health is determined not by one, but by many habits.