Process for selecting microcomputer-based statistical software

Abstract

The proliferation of microcomputers in academic environments has increased interest in using them for statistical computations. In the past many of these computations were done with mainframe or minicomputers using a limited number of highly sophisticated and well supported packages. With the current influx of microcomputer based packages, faculty and researchers are moving their statistical applications from the larger systems to microcomputers. In addition, many faculty and researchers, unfamiliar with large computer systems, are rapidly discovering the use of microcomputer for statistical analysis.The Instruction and Research Computer Center at The Ohio State University has over the last four to five years tried to take a leadership role in the selection of microcomputer software. Basically, the Center assessed the need for software and match that need with the best software for the University. Approximately one and a half years ago the Center began evaluating microcomputer statistical software. In the past, evaluation teams usually consisted ordinarily of the Director of the Center and our microcomputer experts. This approach worked extremely well for communications, basic wordprocessing, basic spreadsheets, and simple databases. The variety of available statistical software packages and the diverse needs of our faculty and researchers dictated a different approach. A statistical software review committee evolved. This committee consisted of faculty, researchers, the center's microcomputer experts, the associate director, and the director.This document reviews the steps taken at The Ohio State University to select the software. We are only partially through the analysis and selection process. Hopefully by the time of the conference, we will able to report about the completed project.