Nonmathematical Statistics: A New Direction for the Undergraduate Discipline

Article excerpt


By the undergraduate statistics discipline I mean those things in teaching, research, and service that are necessary to sustain an undergraduate program (degree program or area of concentration) in statistics. In particular I have chosen to focus on the curriculum for undergraduate majors. I have excluded from my discussion service courses where many positive developments have taken place in recent years. I have also excluded graduate degree programs although some of my remarks might apply there as well.

Many have characterized this as the age of information. The collection, management, and analysis of data are essential to today's society. The things that are at the core of the discipline of statistics are things that society demands. In spite of this, statistics is not widely established as an undergraduate discipline.

I think it is safe to say that most statistics departments began with a central mission to produce M.S. and Ph.D. statisticians to fill needs in academia, government, and industry. A browse through issues of The American Statistician from the 1950s will show great concern for the shortage of statisticians with advanced degrees who would be capable of teaching statistics in colleges and universities and filling research positions in industry and government. An article entitled "Demand And Supply of Statisticians Now and in the Future," which consists of three short papers by Bancroft, Futransky, and Harshbarger (1958), sheds light on the extreme shortage of statisticians at that time. Perhaps because of the need to train those with advanced degrees in statistics, undergraduate programs have generally not been regarded as being as important to the profession as graduate programs.

There are around 220 universities and colleges that offer programs in statistics. This constitutes only about 10% of the four-year colleges and universities in the country. Approximately 60% of these programs are in nonstatistics departments or departments that combine statistics with some other discipline, principally mathematics. The programs that are in statistics or biostatistics departments are located mostly at Ph.D. granting institutions. Statistics programs at four-year liberal arts colleges are virtually nonexistent. By contrast over 50% of the colleges and universities have programs in computer science, and many have separate departments of computer science at the undergraduate level (Barron's 1995).

While statistical reasoning may be considered an important component of a liberal education, statistics is not regarded as a fundamental undergraduate discipline in the same sense that mathematics, history, and biology are. This has had negative consequences not only for the statistics profession but also for society which may not be getting the best trained individuals to handle data collection and analysis. Minton (1983) dealt in detail with the lack of visibility of statistics and concluded that the only way to overcome this problem "is to develop undergraduate degree programs and departments."

In times of tight budgets, graduate programs and small undergraduate programs are usually the first to get careful scrutiny for budget cutting. It may make perfectly good sense to an administrator to eliminate statistics in the name of efficiency or combine it with mathematics or some other discipline. The article "Statistics Departments Under Siege" by Iman (1994) was a chilling assessment of the status of our discipline. The article by Kettenring (1997b) - "The Birth, Life, and Death of Statistics Departments" - again raised concerns about the health of the discipline. Instead of being on the rise in the information age, statistics at some places is on the decline.

An oft-prescribed remedy for the perceived ills of the discipline is that we need to do a better job of selling statistics. As I look back at the 28 years that I've been in the profession, I can see a tremendous "sales" effort. …