WesVar uses one of five replication methods to calculate variances of survey estimates. To do so, we first create replicate weights or import existing replicate weights. In Figure 3, we specify a variable that defines strata and another that defines the primary sampling units (PSUs). We also select a method of variance estimation. In this example, the sample design had two PSUs in each stratum. Balanced repeated replication (BRR) was specified, and WesVar automatically selected a Hadamard matrix of appropriate size based on the number of strata in the data file.
Figure 3. Specify the variables that identify stratum (VarStrat) and PSU (VarUnit). The BRR method of variance estimation is selected.
In Figure 4, replicate weights have been created and are labeled RPL01-RPL24. The user can also type in a different prefix in the Create Weights screen.
Figure 4. Replicate weights were created using the BRR method. From this screen the user saves the WesVar data file. In this case, the name Demo.var was used.
 |
| |  |
About WesVar |
Download |
Product Specifications |
Quick Steps to WesVarTechnical Papers | About Westat