SPSS has just released a new version, 17.0. There are a number of new features, but the most major one for most researchers is that it now offers Multiple Imputation of Missing Values for both categorical and continuous variables. This feature, however, is part of the Missing Values add-on module. If your site license doesn’t have it, ask for it. It’s worth it.
Multiple Imputation has been around as a technique for 10-15 years, but at first it was available only in specialized software such as Sudaan. One by one the major statistical packages have made it available, such as SAS and Splus. But SPSS users have been left behind until now.
If you’re not familiar with multiple imputation, it is one of a few modern techniques for dealing with missing data. All have the advantages of using the full sample size of the data set and creating unbiased parameter estimates. These are both huge improvements over traditional techniques for dealing with missing data, such as listwise deletion or single imputation.
The full list of new features in SPSS 17.0 is available at SPSS’s website, including feature comparisons across versions.