Model World: The Spreadsheeting of OR/MS

Abstract

Spreadsheets are probably the most im portant analytical aid for business since the development of COBOL. Except for the difficulties in debugging, verifying, updating, and documenting spreadsheets, we have no complaint with their abilities and general use. We are fans of spread sheets. We use them all the time. How ever, some issues relative to their role as a pedagogical and analysis medium in OR/MS are of concern: (1) the current touting and overreliance on spreadsheets in teaching OR/MS courses; (2) the trend to incorporate just about all OR/MS meth ods as spreadsheet add-ins, implying that OR/MS analysis and teaching should be done only with spreadsheets; and, from a broader perspective, (3) the use of linear programming (LP) spreadsheet solver add-ins instead of stand-alone LP software for solving classroom and real-world problems. This article is one of a continuing, if in frequent, series of opinion columns. We take strong positions, some backed up by fact and others by our collective judgment and opinion. We are motivated by our concerns with OR/MS pedagogical and practice issues. We feel that an airing of our contrary view is long overdue, given the many INFORMS-published articles that present the view from the other side. In February 1996, the operating subcom mittee of the INFORMS Business School Education Task Force reported on the re sults of a survey of 306 faculty members who taught OR/MS in business schools (2,200 faculty were surveyed with a re sponse rate of 14 percent). The survey showed that ?31 percent of respondents had students use spreadsheets to model and solve