A Few Good Statistical Analysis Tools
Whether your organization needs to measure its programmatic success or expand its donor base, statistical analysis software can help nonprofits become more data-driven and accountable.
If this sounds appealing, you’re in luck: powerful tools that only five or ten years ago required dedicated servers to run can now be installed on a desktop. But these packages can require a considerable investment in up-front and ongoing costs, staff time, and training. Overwhelmed by the choices and not sure if you have the necessary statistical chops? We talked to a statistician to get the latest on some of the options available.
Low-Cost Options and Open-Source Tools
If your software budget is tight and your statistical analysis needs are basic, consider Excel. You probably have it and understand the platform. Drawbacks: Excel can’t automatically handle missing cell values, and its statistical output can appear clunky and scattered. Windows users can install the built-in Analysis Toolpak, but Excel for the Mac OS now requires a free downloadable add-on, like Statplus: Mac LE. These tools amp up your spreadsheet software with features like regression analysis, analysis of variance, and sampling. Excel costs about $120 per license as a stand-alone product, with volume discounts available.
The most popular open-source statistical software, R requires some programming knowledge to navigate its command-line interface. Users enter lines of code to execute R’s functions, but even those lacking a sophisticated computer science background can learn it quickly. R runs on a variety of operating systems, and its thriving user community will help if you get stuck. Nonprofit staffers familiar with programming basics and with a firm grasp of statistical concepts may find R a good choice.
For More Advanced Needs
For users with moderate-sized data sets, Stata is an affordable option, starting at $1,195 ($600 for academic users) for STATA/IC, the standard version. The software lacks the power of some other options on the market, and can only open one data set at a time. Unlike other proprietary software, though, Stata is easily customized, and can handle downloadable user-written commands that significantly expand its capabilities. Stata also draws praise for its tech support, helpful user community, and relative ease of use. Stata has both a graphic, menu-based interface and a command-line interface for those with more programming know-how.
IBM's answer to statistical analysis receives high marks for its user-friendliness. In addition to a syntax editor, SPSS Statistics has a point-and-click graphical interface that doesn't require substantial programming knowledge. This ease of use comes at the expense of some control over statistical output. Nonprofits in need of basic statistical analysis won't find this an issue, but if you seek to do more sophisticated data manipulation, SPSS might prove frustrating. The standard desktop package starts at $5,120 for a user license and a year of support, with higher pricing for concurrent use.
The Top of the Line
With more than one-third of the market share, the SAS Institute is the giant of the statistical analysis software scene. Strengths include power and efficiency in linking large data sets, and a comprehensive built-in set of statistical analysis features. SAS Analytics Pro, the entry-level desktop version of the software, costs $8,500 per user for first-year license fees alone and about $2,000 per year for ongoing use. This software is not for novices, and requires a high degree of statistical and technological expertise to run it. However, SAS offers excellent tech support, and its prevalence means finding others in your network who use the software will be a snap.
A Note of Caution
If your statistical background consists solely of hazy memories of Stats 101, a refresher might be in order. Misrepresenting data, even if done unintentionally, will get a nonprofit into trouble with its donors, board, and other stakeholders. In this case, a little information can be a bad thing.
Thanks to Henry Quinn for his recommendations and advice.
