In the short history of this column, I’ve summarized findings from statistical surveys relating to student IT use at the University of Maryland. In order to analyze these (and other) data, I made use of some powerful computing tools. Here I will describe some of the tools and services available through the Office of Information Technology (OIT) that could help you squeeze information out of your own data.

This Summer, nearly 24,000 course evaluations for Spring courses were scanned and summarized by OIT. That's every course in Architecture, CMPS, and Engineering, as well as numerous other programs and departments. Merely getting this much data into the computer is a daunting task. Until the day when students have personal web devices in every classroom, the tried-and-true optical scan forms (use a number 2 pencil, please!) will remain a practical choice. OIT provides scanning for evaluations, as well as test scoring and surveys. Visit http://www.oit.umd.edu/tel/opscan.html or contact Rose Madison at 301.405.2935.

Scanning is only one step toward taming data. The Statistics Laboratory (http://www.oit.umd.edu/statlab) created the programs for summarizing these course evaluations, customized to the needs of each academic group. More generally, the Stat Lab can help with data processing for evaluations, program assessment, and other needs critical to units at the university.

SAS and SPSS are two of the most widely used software for data summary and analysis. They have been the tools of choice for almost all of the analysis projects in the Stat Lab, including the IT surveys described in previous columns. Both are high-level programming languages complete with toolkits of handy routines such as frequency tables, means, correlations, and many, many more specialized statistical methods. The OIT Software Licensing Office (http://www.oit.umd.edu/slic/) makes them available at very favorable educational discounts. SAS is available on the WAM, Glue, and OIT cluster Unix environments. In addition, both SAS and SPSS for Windows are available in WAM labs.

The larger focus of the Stat Lab is to provide data analysis support for academic research. This includes specialized programming for statistical computing, but it also encompasses other dimensions of good statistical practice. One of the best ways to use the Stat Lab is in the planning stages of a research project or grant proposal. The Lab can help determine appropriate designs and sample sizes, develop effective survey questionnaires, suggest techniques of data coding, and—often most importantly—serve as an objective critic to ask tough questions about the project while there’s still time to make course corrections.

The Visualization and Presentation Lab (VPL, http://www.vpl.umd.edu) is another resource for data-driven projects. The VPL provides expertise and technology to turn research findings into effective presentations. Researchers across the university communicate their work with the help of the VPL’s wide-format color printing, 3-D video, and other hardware and software designed for the visual representation of information. This past Spring, students in College Park Scholars worked with the VPL to produce conference-quality posters for Undergraduate Research Day. (See VPL story.) Visualization is not just a pretty picture; the staff in the VPL can work with researchers to develop and apply new ways of understanding data—especially, large, multi-dimensional data—through computation-intensive visualization.

Open a New Window to Rate This Article