Systematic Sampling 4 Systematic Sampling.pptVIP

Chapter4Systematic Sampling 4.1 Procedures and Uses of Systematic Selection 4.1.A Uses of Systematic Samples concept : Systematic sampling denotes the selection of sampling units in sequences separated in lists by the interval of selection. Briefly, it consists of taking every kth sampling unit after a random start. “pseudo-random selection” 4.1B Problems with Intervals Example2: if the population is made up of 10 elements , which is Yi (i=1,2,3, …10) ,if the K=3, then the three samples are as follow: 4.Using fractional intervals is simple with a decimal fraction . N=920 n=100 k=N/n =9.2 9.2*10=92 Select a random start from 1 to 92 ,then add the interval 92 successively , then round down these numbers by eliminating the last digits. Have a try ！ A population consists of 33 elements，the numbers are listed as follow: 1.2.3……..33 how do you get a sample of 6 elements with systematic sampling ? 4.2 Systematic Sampling of Different Populations Eg4: the population is consisted of the figures of 1、2、3、4、5、6、7、8、9、10、11、12, if we want to choose 4 figures with systematic , please compare the efficiencies of random SRS and ordered sample. 4.2 Systematic Sampling of Different Populations 1. Random Ordered Population (the elements’ arrangement is unrelated with Yi) common procedures of systematic sampling In fact , there are N! different arrangements of the elements, the expectation of the efficiency is the same to the efficiency of SRS. 2. A population with monotonic trend The mean of the systematic sample depends greatly on the selection of the random start. The monotonic trend leads to large differences among all possible sample means. ◇when N=nk and n is odd ，the first n-1 parts are drawn in the same way ，but for the last part，we should choose the element of [r+(n-1)k]，that is the rth element of the last part. When N=nk and n is odd the sample contains (n-1)/2 pairs of elements and another element .