标题:
高校学生安全事故发生状况及其影响因素的统计分析
The Statistical Analysis about Status and
Influencing Factors of University Students’ Safety Accidents
作者:
黄希芬, 何伟全, 化存才, 尚云
关键字:
校园安全, 影响因素, Logistic回归, 交叉表分析Campus Safety, Influencing Factors, Logistic Regression, Analysis of Crosstab
期刊名称:
《Statistics and Application》, Vol.3 No.2, 2014-06-10
摘要:
校园安全是高校学生管理中长期关注的问题，是高校网络舆情爆发的重要诱因，深入研究高校学生安全事故的发生状况及其影响因素，能为有效预防、监控和干预大学生伤亡事故提供科学的依据。为了掌握高校学生安全事故发生状况与其发生的影响因素，本文根据云南某高校所记录的大学生伤亡事故统计数据表，并设计统一的调查问卷，对云南省某高校的210名在校学生进行高校学生安全事故情况调查。首先，采用交叉表方法对调查数据进行了简单的统计分析；其次，采用单因素进行筛选，再用多因素Logistic回归分析方法研究安全事故发生的主要影响因素及其影响力度，得到的研究结果表明：“饮酒”、“心理状况”和“宿舍关系”是影响该高校安全事故发生的主要因素，其影响程度按“心理状况”、“饮酒”、“宿舍关系”依次递减；最后，通过现有的数据对建立的Logistic回归模型的预测准确率进行检验，得出模型的总正判率达到92.9%，说明模型的预测较准确。
Campus safety is the issue of long-term focus on students’ management in colleges and universities. It is an important cause of the outbreak of network public opinion. Further study of safety accidents’ situation and its influencing factors of college students could provide scientific basis for effective prevention, monitoring and intervention of college accidents. In order to grasp the safety accidents’ happening situation of college students and its influencing factors, we design a questionnaire and investigate 210 students about their situation on campus accidents in a collegeofYunnanprovince according to the casualty statistics table of some college students of the college. First of all, we apply the statistical analysis of crosstab method to present some statistical information. Secondly, we apply the single factor analysis and the multi-factor Logistic regression analysis method to explore the main influencing factors of safety accidents and their impact degrees. Our results show that the “drinking”, “psychological situation” and “dormitory relationship” are the main factors influencing the safety accidents in the college, and the influence degrees decrease with respective to the “psychological situation”, “drinking” and “dormitory relationship”. Finally, we test the prediction accuracy of the established Logistic regression model using the existing data and find that the total rate of the model is sentenced to rate of 92.9%. This implies that the model’s prediction is more accurate.