For analysts who want a single package to perform a variety of statistical analyses, STATA provides complete management, graphing and modeling for large data sets. Unlike more streamlined software packages, STATA has a user friendly interface; while the command line takes some getting used to, the learning curve is relatively smooth for those who have worked with related programs such as SAS and Matlab. Overall, STATA is a good tool for business analysts and researchers who are looking to run a variety of regression techniques on large data sets.
The package we tested is STATA MP, which allows for over 10,000 independent variables across any number of observations; while we ran the PC version, both Mac and Linux users can run the software natively as well. For smaller data sets, the IC edition provides a budget version which is limited to just over 2,000 independent variables (a trial “Small” version provides up to 99 variables across 1,000 observations.) While the software can be a bit expensive – close to $1,000 for a single-user license – I found the built-in commands and options to save a lot of time relative to less structured packages.
Within minutes, you can load data in from a variety of sources (XML, CSV or virtually any database) and begin performing commands, with integrated options for regression, survey sampling models and, of course, summary statistics. Version 10 has much improved graphing capabilities, allowing you to produce clean, concise reports for research or professional papers. STATA provides on-going support to their users by way of on-line manuals, technical support engineers and training seminars – all of which make it a good community to join as an analyst.
