Often I have to answer the question about what software to use for statistical analysis. The answer is not that straightforward. It depends primarily upon your analytical needs. I recommend the following software based on your needs:
- If you are conducting basic analytics with a small data set, use Microsoft Excel. You can use the built in Analytic tool box (not available in the latest version for Mac) to do descriptive statistics, correlations, regression models, and other statistical tests.
- If your analysis goes beyond the basic regression models where you may have to estimate models using maximum likelihood routines, use SPSS. It is easy to learn and is fairly powerful for undergraduate level research needs. (www.spss.com)
- If you need to write custom code, such as writing user defined maximum likelihood functions, I would recommend Stata, which is very similar to SPSS, but a whole lot powerful. Stata is sufficient for even doctoral level research in Econometrics while continuing to be fairly easy to learn. (www.stata.com)
- SAS is another option for advanced analytics. However, SAS is the tool of choice for the older generations. SAS has reluctantly embraced the changing computing platforms and therefore has the look and feel of a software from 70’s. SAS is very powerful, but least bit Sassy! (www.sas.com).
- My favorite statistical software is R, which is a freeware. R is fast becoming popular, especially after John Fox of McMaster university developed the GUI for R giving it the point and click capabilities. R is similar to Stata. There are hundreds, if not thousands, of researchers developing advanced tools for R and making them available from R’s website. The number of R add-ins exceeded 2,000 in October 2009. Lastly, a new book in 2009, R Through Excel, allows R to be run almost seamlessly from within MS Excel. The two best features of R are:
- It is absolutely free, no strings attached.
- It is extremely flexible for any advanced research in statistical methods.
A quick view of internet site traffic suggests that SAS continues to lead the market share in statistical software. However, the graph below suggests that SAS is fast loosing the market’s interest where the daily traffic to its site collapsed from almost 30,000 unique visits in July 2007 to 15,000 in August 2009. Even though R is a freeware, it is attracting more traffic to its website than the other commercial vendors, i.e., SPSS and Stata.
I have created a channel on YouTube to post training videos using R, SPSS, and Excel. In December, I will be uploading 20-hours of videos on a course in statistical methods and research.