Why is it that statistics generally seems interesting, even exciting, to the practitioner while the same subject seems dull and irrelevant to the student? Could it be that there is a large gap between how statistics is practiced and how it is taught? Today, the practice of statistics is open and dynamic, no longer constrained by narrow mathematical formalism and tight rules of inference. In this mode, it has become key to industrial revitalization on one hand and the revitalization of K-12 mathematics programs on the other. Colleges and universities, caught in the middle, are only now beginning to replace the classical formalism with a more vibrant approach. The Activity- Based Statistics (ABS) Project is an attempt to bring the most successful aspects of the industrial uses and teaching of statistics and of the K-12 mathematics curriculum reform to bear on the content and style of the introductory college course. This is to be accomplished with an active approach that engages the student's mind, hands, and eyes in the "discovery" of the concepts of statistics followed by the application of these concepts to significant problems from the student's world.
Current practice of statistics involves much exploratory analysis to focus the question or problem under investigation, careful planning of data collection schemes to answer specific questions, and some inferential (confirmatory) analysis to frame proper conclusions. The weight is on exploring, planning and informal inference, with probability actually playing a rather minor role. This runs counter to the outline for most introductory courses, and therein lies a problem. The ABS Project will present a structure for introductory statistics that is more in line with statistical practice, and provide modules for students to learn basic concepts while engaged in directed activities.
Activities will be organized around five major themes: process improvement (using the philosophy and tools of TQM), sample surveys, experiments, observational studies, and modeling. After some introductory notes (the material is designed to supplement a textbook, not replace it), activities for students will illustrate the concepts and tools involved in each area, such as the use of control charts, the role of randomization in sample surveys or experiments, the notion of factorial arrangements of treatments, and the regression effect. The statistical concepts and tools are introduced as needed to solve a problem, not simply as a list of things to be learned. Statistics is thought of as problem solving in a broad sense.
Drafts of many of the activities envisioned for the final product have been field tested in statistics classrooms, but more field testers are needed. Anyone interested in field testing activities should contact Richard Scheaffer.
Ideas for new activities are always welcome!