This study aimed at examining the differences in emotional, social, and moral characteristics of adolescent students who showed different amount of violent behavior while aggregating in group against outsiders. There were 4 research objectives, First, to examin the differences in moral ressoning ability, future orientation trait, mental health and attitudes towards violence in various types of violent students classified by age, socioeconomic status of the family, major subjects of study, and academic achievement. Secondly, to study the relationships among and magnitude of violent behavior, desirable peer group experience and democratic atmosphere among teachers of vocational students. Thirdly, to compare and contrast the predictive power of a set of psychological traits and a set of reported 
social conditions used alone or combined as accounting for the different amount of violent behavior of these students. Finally, to identify the students who behaved with different degrees of violent, according to the bio-social characteristics of the student. 
                The sample in this study consisted of 250 male adolescent students from 4 vocational colleges in Bangkok. Their ages ranged from 13 to 23 years, with the average age of 17.20 years. They were studing in one of the 3 major subject areas namely, mechanical, electrical and  constructional areas with the total of 250 students. 
                In this field study, the questionnaires booklet consisted of three parts; 
(1) the assessments of the bio-social characteristics of the respondents and the amount of verbal and behavioral violence engaged in while being enrolled in college. (2) the assessment of the 4 psycholgical traits which were moral reasoning ability, future orientation  trait, mental health and attitudes towards violence. (3) the assessments of 2 perceived social conditions in the college which were desirable peer group experience and democratic atmosphere among teachers and students. The psychological and social  variables were assessed by 10 to 24 item tests, each was accompanied by 6- unit rating scale. 
                Four hypotheses were tested in this study by carrying out the three-way Analysis of Variance and the Stepwised Multiple Regression Analysis, and the intercorrelation Matrices on the data of the total sample and the subsamples. 
           There were 6 major findings: 
               First, the male students under studied who had never participated in violent event numbered 192 students (76.80 percents of the sample) On the other hand, the students who engaged in violent behavior more often were, older students, upperclassmen students and at the same time showed higher academic achievement, especially among the construction students, and students from higher socioeconomic families in general. 
           Secondly, it was found that the less violent students or the nonviolent students held less favorable attitudes towards violenct behaviors than their counterparts. This attitude was the best predictor of violent behavior in most students except the mechanics students. 
           Thirdly, negative attitudes towards violence was found to be highly related to favorable peer group experience. Fourthly the students with less violent behavior were found to highly perceive the democratic atmosphere in the college more than their counterparts. In turn, the perception of democratic atmosphere was found to be positively related to the desirable peer group experience of these students. 
           Fifthly, it was found that the future orientation scores of these students were related to negative attitude scores, perceived democratic atmosphere scores, favorable peer experience score, positive mental health scores and high moral reasoning ability scores of these male students. 
           Finally, the 4 psychological traits together could better predict the magnitude of violent behavior of these studets than the set of two social variables. However, when combined these 2 sets of predictors, this new set of predictors could better predict the violent  behavior of the 3 types of students namely, high academic achievers, younger students and construction students. 
                From these research findings several suggestions were made as followed. For the prevention of violence among vocational students of different colleges, favorable attitudes towards nonviolent behavior,  more appropriate friendship experience, future orientation, perception of democratic atmosphere in the institution and good mental health should be promoted, Furthermore, special attention from teachers and college administrators should be given to the following types of students: older students, students from well-to-do families and students who received good grades in general and especially among the construction students, in order that the present and future problems can be solved.