Statistical Method And Application For Two Quantitative Data RRT Models Under Stratified Two-stage Cluster Sampling

Objective:In this thesis, we combined additive model and multiply model of quantitativequestion with stratified two-stage cluster sampling and derived the relevant formulae,which were used in the survey of MSM populations in Beijing about their quantitativesensitive questions. In this way, we provide effective and reliable method to futuresensitive question survey under stratified two-stage cluster sampling in larger scale.Methods:Basic theorem and methods of Probability Theory and mathematical statistics suchas Cochran’s classic sampling theories, Total Probability Theorem, characteristic ofmean and variance were applied though out the formulae deriving. Stratified two-stagecluster sampling method was adopted when investigating quantitative sensitivecharacters of MSM populations in Beijing. Before investigation, we divide Beijing intourban and rural stratification. In the first stage of sampling, three districts, totally sixdistricts were respectively and randomly extracted from two stratifications. In thesecond stage of sampling, totally twenty-three survey points were randomly extractedfrom the six districts. Then we investigated the MSM individuals by additive model ofRRT in all of the extracted points and used the formulae we derived to calculate theestimator and variance of MSM populations in Beijing.Results:1. The formulae of estimators and their estimated variances of additive modelcombined with stratified two-stage cluster sampling were successfully conducted in this thesis.2. The formulae of estimators and their estimated variances of multiply modelcombined with stratified two-stage cluster sampling were successfully conducted in thisthesis.3. The quantitative sensitive characters and their95%confidence intervals of MSMpopulations in Beijing are as follows:①The average first sexual intercourse age of MSM populations in Beijing is21.36,with the95%confidence interval of (20.37,22.35).②The number of average sexual partners of MSM populations in Beijing was2.64,with the95%confidence interval of (2.20,3.08).③There were4.49times of sexual intercourse for MSM populations in Beijing onaverage, with the95%confidence interval of (3.96，5.02).Conclusions:1. The method and formulae we provide have great application value, by which wesuccessfully investigated the quantitative sensitive characters of MSM in Beijing.2. The survey results indicate that MSM populations in Beijing have the riskysexual behavior of young age of first sexual intercourse, multiple sexual partners,requiring aimed interventions.