The Effect of Nonresponse Primary Sampling Units on Estimating the Variance of Changes by Jackknife Method (Case Study: Labor Force Survey Data for 2009 and 2010)

Abstract. According to the importance of presenting change estimation of labor force survey indicators along with their variance, in this paper, the use of Jackknife method in estimating variance of changes has been investigated. Then, the effect of nonresponse primary sampling units on estimating the variance of changes has been studied by use of Jackknife method via intensive simulation study. Further that, the effect of nonresponse on estimating the variance of key estimations changes has been investigated based on labor force survey data of 2009 and 2010. The results of this study show that if a pseudo-district is nonresponse from a specific point of time in consecutive survey periods, this will affect the estimation of variance of changes of key estimations, however this effect may be small.