Undergraduate Programs and the Future of Academic Statistics

Article excerpt

1. INTRODUCTION

The American Statistical Association is preparing recommendations concerning undergraduate programs in statistics. The proposal that ASA offer its blessings to undergraduate programs is not universally accepted, because statisticians are generally expected to have graduate training. This article does not discuss the content of undergraduate programs, but rather presents some factual background material that we should consider in designing them, with some opinions based on this background. That my title alludes to "the future of academic statistics" suggests the importance I attach to revising our approach to undergraduate statistics.

My themes in summary are: Although the discipline of statistics is healthy, its place in academe is not. Our future there depends strongly on achieving a more prominent place in undergraduate education beyond the first methods course. To this end, we must offer undergraduate programs that are popular with students and suited to institutions that lack statistics departments. In my opinion, the primary intent of such programs cannot be to prepare students for graduate study in statistics, but to equip them for employment with a bachelor's degree or for further study in a wide variety of areas. Finally, success requires greater cooperation between statistics and other disciplines, especially mathematics.

This article has four main sections. Section 2 summarizes the state of academic mathematics and statistics and argues that, in most institutions, the two disciplines need each other. It would repeat false starts from the past to think primarily of statistics departments, or even of large research universities more generally. I believe that growth of undergraduate statistics programs in other institutions will generally require the cooperation of the mathematics department, and that mathematics may be ready for more cooperation. Section 3 presents some market research--data on trends that ought to influence our thinking about statistics for undergraduates. Section 4 offers some cautionary findings from research in mathematics education. The unifying theme of these three sections is the need for realism in discussing programs for undergraduates. It is easy to be unrealistic. The traditional approach--starting by formulating a list of topics that everyone applying statistics should know--is almost certain to gen erate recommendations that most institutions cannot follow and that most students cannot master. The concluding section offers some opinions and advice on the basis of the more factual content in the first three sections.

2. MATHEMATICS AND STATISTICS

The Bureau of Labor Statistics (2000) said that "In 1998, approximately 110 universities offered a master's degree program in statistics, and about 60 offered a doctoral degree program." Table 1, in contrast, shows the universe of degree-granting institutions in the United States (NCES 1999b):

The two-year college totals are impressive, but to begin discussion of undergraduate programs in statistics we might focus on the four-year colleges: more than 2,300 institutions with (by now) more than 7 million undergraduate students. Few of these institutions have statistics departments. If we wish to encourage study of statistics by undergraduates, we shall have to gain the cooperation of other disciplines. In the case of undergraduate majors, the mathematics department is the natural home of statistics. Minors or concentrations in statistics may target students in other disciplines, but even these will often require the cooperation of the mathematics department. Of course, the mathematics departments at many institutions lack the resources to offer more than a few statistics courses. Realistic plans will keep this in mind.

Moore and Cobb (2000), in a forum read by mathematicians, discussed in detail the need for cooperation between mathematics and academic statistics. …