(From Amstat News, November 1990, pp. 19-20)

The University of Iowa

Thirty-nine statisticians -- representing universities, colleges, consulting firms, business and industry -- gathered for a workshop on statistical education in Iowa City, Iowa, June 18-20, 1990. The workshop was sponsored by the University of Iowa and ASA with financial assistance provided by the National Science Foundation, the ALCOA Foundation, the Ott Foundation, the Statistics Division of the American Society of Quality Control, and the Quality and Productivity Section of ASA.

As a group, we recognized several poor characteristics of science and mathematical education, including statistical education.

* Improvements are needed in the K-12 curriculum, as there is widespread science and mathematics illiteracy in the United States.

* There is little effort given to recruiting students to these areas, often because we want only students "like ourselves."

* There is a lack of truly qualified teachers because the pool from which they come is drying up.

* College and university instructors are often required to teach large introductory classes allowing little or no interaction with students and severely limiting student involvement with others.

* Students frequently view courses as being "hard" because the class periods are dull and it is difficult to get good grades despite spending long hours doing homework.

* Sometimes there is a communication barrier. The instructors often forget to emphasize the "big picture" by concentrating on the solutions of the problems under consideration and eliminating any in-depth discussions of the basic ideas.

* In most universities and some colleges, there is no sense of community among most students in these areas. Accordingly, there is not much discussion during and after class periods. Somehow we should encourage teamwork and stress the importance of it.

Many of us believe that those of us in universities must "reach out" more to a number of groups, including the following.

* The high schools, in particular working with future and present high school teachers.

* Colleges, with much more interaction between university and college professors and students.

* Graduate students, working with them on their teaching as well as their research.

* Junior faculty, serving as mentors in teaching as well as research.

* Colleagues in other fields, interacting and working on joint research projects.

* Business/industry/government, because many more joint efforts are needed here if America is going to continue to be in the "first class" category.

* The public, trying to improve numeracy and science literacy.

However, it is time consuming to participate in these activities; thus we must "change the present system" if our faculties are to get credit for these efforts.

Clearly in a three-day workshop we could not consider all of the problems listed above. We thought that we could do the most good by addressing the problems associated with our introductory courses, some of which are:

1. Statistics teaching is often stagnant; statistics teachers resist change. The most popular elementary texts evolve but slowly over decades. Meanwhile statistics is progressing rapidly.

2. Techniques are often taught in isolation with inadequate motivation and with no connection to the philosophy that connects them to real events; students often fail to see the personal relevance of statistics because interesting and relevant applications are rare in many statistics courses. The open-ended nature of statistical investigations and the sequential nature of statistical inquiry are not brought out.

3. Statistics is too often presented as a branch of mathematics, and good statistics is often equated with mathematical rigor or purity, rather than with careful thinking.

4. Teachers are often unimaginative in their methods of delivery, relying almost exclusively on traditional lecture/discussion. They fail to take into account the different ways in which different students may learn, both individually and in groups, or the many possible modalities of teaching. Often "megaclasses" cause these problems.

5. There is little attempt to measure what statistics courses accomplish. Statistics is too little used.

6. Many teachers have inadequate backgrounds, both in knowledge of the subject and ability to communicate in English. The word "statistics" has itself acquired bad connotations.

7. Statisticians may put their subject in a bad light for the students. They often fail to see any need to convey a sense of excitement.

8. Some teachers are technically incompetent, either in aspects of statistics or in the underlying mathematical tools. They may mislead by treating statistical investigations as if they entailed random sampling from some finite population.

Some of the major suggestions for these courses concerned the following:

* State the goal(s)

* Analyze data and do projects

* Most should use the computer

* Lecture less, teach more

We also listed topics that seem important for a first course (1-4 highest priority, 5-9 second, 10-13 third, 14-17 fourth).

1. Recognizing that statistics is everywhere.

2. Understanding variability.

3. Collection and summarization of data.

4. Graphs, including time series plots.

5. Sampling and surveys.

6. Elementary designs of experiments.

7. Formulation of problems.

8. Basic distributions.

9. Correlation and regression and other measures of association.

10. Elementary probability.

11. Central limit theorem and law of averages.

12. Elementary inference from samples.

13. Ability to use at least one statistical software package.

14. Dealing with outliers.

15. Statistical significance vs. practical significance.

16. Categorical data and contingency tables.

17. Simulation.

Finally we suggested items for ASA (or the statistical profession) to consider to help improve the teaching of statistics.

* Construction of something like MAA's The College Mathematics Journal.

* Newsletter on statistical education.

* Providing information on teaching TAs to teach (MAA has one which was edited by Betty Ann Case). Encourage extensive TA training and mentoring.

* Developing teachers' network.

* Workshops on teaching at ASA meetings.

* Short courses on teaching.

* Poster sessions on teaching, student projects, and data sets.

* Supporting faculty efforts to change the system.

* Diagnostic tests for students.

* Encourage position papers, each prepared by two or three like-minded persons.

* Funding for future conferences or workshops on statistical education.

* Collection of "tidbits", namely good ideas for classroom teaching.

* Writing a "Point of View" article for last page of The Chronicle of Higher Education.

* Supporting efforts to modify the academic system, in particular the reward structure and grading procedure.

* Becoming aware of difference of jargon. For example, an engineer might call the statistician's independent variable (or factor) a "parameter."

* Center for collecting materials like projects, videos and case studies.

Look for a complete report on the workshop in an upcoming edition of The American Statistician.

Return to Cobb Paper | Return to Table of Contents | Return to the JSE Home Page