Regression analysis of Complex Sample Survey Data: results from a national Survey

Abstract

Abstract. This study made comparison if there are differences in mean and significant probability depending on application method of weights when chisquare test is performed using Complex Sample Survey Data. Multiple regression and complex sample regression analysis were used as analyzing methods. Youth Risk Behavior Web-based Survey (KYRBS) 2012 was used as data source. As the result of multiple regression and complex sample regression analysis, there were differences in mean and significant probability. The result of this study confirmed that in the analysis of data from complex sample survey, biased variance (standard error) may be drawn out when using Simple Random Sampling Design.

Analysis of Regression Based on Sampling Weights in Complex Sample Survey: Data from the Korea Youth Risk Behavior Web-Based Survey

“…In addition, Lv , et al, [7] also pointed out that when only frequency weight is applied to cluster sampling data, there is a possibility that significance levels of all variables may become smaller as the size of the sample becomes bigger [6]. Nevertheless, many studies do not gain accurate analysis results as they perform statistical analysis , wrongly assuming the data from stratified sampling as those from simple random sampling [5,8,15]. As an alternative to this problem, Lee [16] suggested using the Rao-scott chi-square test as a categorical analysis method for complex sampling data and, by using data from Third National Health and Nutrition Examination Study of the U.S., proved that Rao-scott chi-square test can present more stable statistical figures than the Wald test.…”