## Thursday, December 4, 2014

### CCHS Bootstrap Weights

Question

A researcher is looking for the bootstrap weights for the 2003, 2005, and 2007 CCHS. Are these available? I did find the information in the Nesstar WebView for Cycle 2.1 (2003):

"Weighting - The principle behind estimation in a probability sample is that each person in the sample "represents", besides himself or herself, several other persons not in the sample. For example, in a simple random 2% sample of the population, each person in the sample represents 50 persons in the population. In the terminology used here, it can be said that each person has a weight of 50. The weighting phase is a step that calculates, for each person, his or her associated sampling weight. This weight must be used to derive meaningful estimates from the survey. For example, if the number of individuals who had a major depressive episode is to be estimated, the weights of survey respondents having that characteristic should be summed. In order for estimates produced from survey data to be representative of the covered population and not just the sample itself, a user must incorporate the survey weights into their calculations. In order to determine the quality of an estimate, the variance must be calculated. Because the CCHS uses a multi-stage survey design, there is no simple formula that can be used to calculate variance estimates. Therefore, an approximative method is needed. Coefficient of variation, standard deviation and confidence intervals can then be calculated from the variance. Thebootstrap re-sampling method used in the CCHS involves the selection of simple random samples known as replicates, and the calculation of the variation between the estimates from replicate to replicate. In each stratum, a simple random sample of (n-1) of the n clusters is selected with replacement to form a replicate. Note that since the selection is with replacement, a cluster may be chosen more than once. In each replicate, the survey weight for each record in the (n-1) selected clusters is recalculated. These weights are then post-stratified according to demographic information in the same way as the sampling design weights in order to obtain the final bootstrap weights. The entire process (selecting simple random samples, recalculating and post-stratifying weights for each stratum) is repeated B times, where B is large. The CCHS typically uses B=500, to produce 500 bootstrap weights. To obtain the bootstrap variance estimator, the point estimate for each of the B samples must be calculated. The standard deviation of these estimates is the bootstrap variance estimator. Statistics Canada has developed a program that can perform all of these calculations for the user: the Bootvar program."

Is the Bootvar program what he would need?