Influencing Factors Of Cyberbullying Among College Students In Guizhou Province Of Multilevel Model

Objective:To describe and analyze the prevalence and influencing factors of cyberbullying among college students in Guizhou and to provide a basis for formulating corresponding intervention policies and prevention measures.Methods:The 2669 students from 55 classes in 6 colleges in 3 regions of Guizhou were selected by adopting a multi-stage stratified cluster random sampling method and assessed by using cyberbullying behavior questionnaire and self-made questionnaire on factors affecting cyberbullying.The Cronbach'?coefficient,the split-fold reliability and factor analysis were used to analyze the reliability and validity of the questionnaire.The single factor analysis of the influencing factors used the?~2test and Z test,and multi-factor analysis used traditional single-level Logistic regression model and multi-level Logistic regression model.Results:(1)Cyberbullying behavior questionnaire reliability and validity test:Cronbach'?coefficient is 0.9020,Spearman-Brown split-fold reliability coefficient is 0.8420,Guttman coefficient is0.8400,and average measured ICC value is 0.9020(95%CI:0.8970?0.9080)),the correlation matrix is analyzed by principal component method,and the Kaiser-Meyer-Olkin(KMO)value of this questionnaire is 0.9350 measured by the maximum variance orthogonal rotation measurement,and the approximate chi-square value of the Bartlett sphericity test is 15621.635(P<0.001),suggesting that the questionnaire has good reliability and validity.(2)Cyberbullying detection:Among the 2,669 college students,cyberbullyers were 1728,with detection rates of 64.74%.(3)Analysis of influencing factors of cyberbullying:The results of single factor analysis showed that different genders,ethnicities,family types,family economy,relationship with parents,self-assessment of health,drinking,smoking,number of social software,online time each day,playing online games,internet addiction,academic performance,close friends,student cadres,and college students participating in group activities were statistically significant(P<0.05).The single-level Logistic regression model showed that drinking(OR=1.291,95%CI:1.072?1.553),daily Internet time 4?6h(OR=1.314,95%CI:1.079?1.600),daily Internet time>6h(OR=1.444,95%CI:1.132?1.843),the number of social software>2(OR=1.213,95%CI:1.017?1.447),Internet addiction(OR=1.625,95%CI:1.359?1.942)were the risk factor of cyberbullying;good family economy(OR=0.382,95%CI:0.209?0.699),average relationship with parents(OR=0.327,95%CI:0.154?0.691),good relationship with parents(OR=0.397,95%CI:0.191?0.827),good health self-evaluation(OR=0.780,95%CI:0.612?0.994),student cadres(OR=0.811,95%CI:0.674?0.977),close friends(OR=0.313,95%CI:0.145?0.676)were the protective factor of cyberbullying.The two-level zero model showed that the variance of college students'cyberbullying at the class level(level 2)was statistically significant(?~2=9.006,P<0.05),the data was aggregated at a high level,and the hierarchical structure could not be ignored.The results of two-level Logistic regression model showed that drinking(OR=1.287,95%CI:1.066?1.553),daily online time 4?6h(OR=1.298,95%CI:1.063?1.586),daily online time>6h(OR=1.419,95%CI:1.106?1.820),the number of social software>2(OR=1.218,95%CI:1.017?1.458),Internet addiction(OR=1.576,95%CI:1.314?1.891)were the risk factor of cyberbullying;good family economy(OR=0.363,95%CI:0.197?0.669),average relationship with parents(OR=0.324,95%CI:0.151?0.693),good relationship with parents(OR=0.401,95%CI:0.191?0.842),student cadres(OR=0.823,95%CI:0.682?0.993),close friends(OR=0.293,95%CI:0.134?0.640)were the protective factors of cyberbullying.Conclusion:The cyberbullying behavior questionnaire is suitable for evaluating college students in Guizhou;The college cyberbullying detection rate of college students in Guizhou is high;Cyberbullying of college students in Guizhou is clustered at the class level,indicating that the multi-level model analysis of the factors affecting cyberbullying is reasonable;The occurrence of cyberbullying by college students is related to their personal family situation,class situation,life behavior and network behavior.