| Objective: Premarital sexual behavior may not only causes great injury to both sides of the involved parties, it also can bring great burden to the society, the routine investigation method, which get the information by questionnaire can hardly get reliable and accurate data[1]. Our investigation adopt the randomized response technique to get data of premarital sexual behavior rate in the college students, in addition, we also implored the influential factors of premarital sexual behavior in order to enhance the conditions of the teenager's reproductive health and to afford evidence for the educational and decision-making department's administration.Methods:1 Sample method we adopt the sample method of cluster sampling: regard the 14 classes as 14 clusters (there is 190 students in every cluster on average), then randomly sample 3 classes form them to launch the preliminary investigation.2 Calculate the sample size at first, according to the formula of calculating the size of cluster sampling which introduced by the Huang Ti Qian[2] calculate the number of cluster k which we will investigate; then, on the base of Q-type clustering analysis calculate the sample size when compare the variables between different ; at last, combine the two sample sizes calculate by the two ways above and determine the final sample size and the number of cluster k.3 Technique method 3.1 Adopt "the general conditions questionnaire on the college students" to get all the possible factors which related with the premarital sexual behavior.3.2 Adopt the three unrelated question randomized response model to get the rate of premarital sexual behavior in the college students4 Statistic methods 4.1 Calculate the premarital sexual rate of p in the college students with the help of the randomized response technique.4.2 In the case of the information on every college student's general features which get by questionnaires launch the Q-type clustering analysis and calculate the final clusters of the samples(NCL), i.e. the clusters of the crowd.4.3 Compare different variables among all the clusters, in which compare the sexual ratio,origination and political features with the RC chi-square test, when compare between every two clusters correct the significance level[3]; All the others variables are compared by the method of Kruskal Wallis test, Nemenyi method is used when compare variables between every two clusters, the calculate the data by the statistic software CS2000(Concise Statistics 2000). Results: 1 Sample size 1.1 According to the formula of clusters sampling from finite clusters, calculated that the number of clusters k is 3.1.2 Calculates from the view of comparing every variable among all the clusters, the sample size is 279.1.3 Calculates from the view of comparing each variable between every two clusters, the sample size is 370.1.4 Combines all the calculation results, determine that the final size is 3 clusters (about 570 students).2 According to the investigation result of the randomized response technique calculated that the sexual behavior rate is 13.38 percent, and the SE is 0.0744.3 According to the clustering analysis, grouped all the samples into three clusters, cluster I, cluster II and cluster III.4 Compare variables among all the clusters in all the tests of below the significance level ()is 0.05 , when compare between every two clusters with the RC X2, correct the significance level[3].4.1 The comparison of the college students' basic conditions by the RC chi-square test, we can not regard the originations of the three clusters are different(X2=4.641,p=0.982), but the sexual ratio(X2=160.060,p=0.000) and political features (X2=81.723,p=0.000) are different, combine the comparison results between every two clusters, we can draw the conclusion that in the cluster I the ratio of male is the highest, while the proportion of communist is the least, on the contrary, in the cluster III, the ratio of male is the least and the proportion of communist is the...
|
PDF Full Text Request