Replication-Based Variance Estimation for Analysis of Complex Survey Data
WesVar® uses the robust and flexible approach of replication variance estimation. Replication methods apply to sample designs and estimators from the simple to the most complex. Using one of five methods of replication, users can estimate standard errors for simple estimators like totals or complicated ones like logistic regression parameter estimates.
Easy to Use
WesVar's simple point-and-click interface lets users quickly create weights, specify tables, and define regression models. Coding of procedure statements is unnecessary.
Multiple Methods to Create Weights
Users can create survey weights using three methods of jackknife replication and two versions of balanced repeated replication. Users select the variables that identify strata and primary sampling units and choose a replication method. WesVar creates a set of weights for each replicate subsample. Users can then adjust the weights for nonresponse, post-stratify the weights, or rake weights to control totals.
Analysis requests are organized in workbooks that have an intuitive tree structure. Highlighting different nodes in the tree allows users to do the following
- Select the statistics they want
- Control the appearance of output
- Set the level of confidence intervals
- Specify many other options
Simple dialog boxes let users define complicated functions of estimates in cells of a table and statistics like means, quantiles, and standardized rates.
Regression models are specified with a few mouse clicks or by dragging and dropping variables. Easy-to-use screens let users define customized hypothesis tests and odds ratios.