Newsletter for the Section on Statistical Education
Volume 4, Number 1 (Winter 1998)
When I first walk into my Introductory Statistics courses, I have two main goals: give the students a flavor for the course and introduce them to the practice of statistics through examples. I don't feel giving a simple definition of "statistics" is meaningful, and instead want them to discover what the subject involves through recent examples of uses and misuses. I also want them to know from the beginning that this course won't be only about number crunching, but will also emphasize reading numerical discussions, writing technical arguments, and reasoning statistically. To accomplish these goals, first I present a series of myths to the students and then we discuss some more realistic viewpoints of the course. Then I present a series of misuses, mostly from the news. By having students identify the errors in the arguments, they begin to develop their own statistical intuition and understanding of statistical practice. Finally, to ensure they believe that not all statistics are lies I present examples of a few effective uses.
Examples of Myths Students Have Entering the Course
To give students an idea of what I will expect from them in the course, I discuss a series of myths they may hold entering the course:
I try to emphasize to the students that since computers can now do the numerical calculations so well, our role in studying statistics has changed to being able to tell the computer the right thing to do and then interpreting the output. I also stress that students will need to be able to justify their answers since multiple interpretations are quite possible. (A good example here is two different newspaper headlines based on the same study with opposite implications.) I also try to convey that statistics will not only be important no matter what career they choose, but also just for evaluating information in the newspaper. To study the material, I encourage them to approach it like a foreign language: immersing themselves in the use of the terms, and constantly practicing "talking statistics" with other students. Also, since the students taking my course are so diverse, I encourage them to work with others to share their distinct perspectives. Finally, I tell them my first goal of the course is for them to examine statistics with a critical eye (instead of accepting whatever numbers they hear) and to become intelligent consumers of statistical arguments.
Examples of (Mis)uses of Statistics
I present the following points on overheads and have the class explain to me how they feel about the statements - if they feel they are effective uses of information and convincing arguments. Most of these examples are borrowed from Chapter 3, Section 4 of Statistics: Concepts and Controversies by David Moore, 3rd edition. Moore has an excellent discussion of how to "look at data intelligently". I also try to accompany these points with recent newspaper headlines illustrating the same misuses.
Discussing this list enables students to develop a list of questions to ask when evaluating any numerical argument: What comparisons are being made? Is the information complete? Are the numbers plausible? Are the definitions clear? Are the right numbers being looked at? What is the source of the information? How was the sample selected? Does the conclusion follow? Are the generalizations valid? I remind them how even the experts have made some serious mistakes (Challenger Accident, Dewey Defeats Truman) and such misuses of statistics should not be taken lightly. Good Uses of Statistics To reassure the students that there are plenty of good uses of statistics, I cite some recent studies that are quite informative. I also share some projects I've been involved with (motivated by Bentley, D. (1994), "My First Day's Lectures: Past and Present" presented at Joint Statistical Meetings) to show them how accessible the questions are to them and the broad variability of disciplines asking the questions. For example, I tell students about my recent work with projects involving
I have found these ideas effective at motivating student interest in the course from day one, as well as aligning their expectations of the course with my own. These ideas are reinforced by their first homework, to identify and evaluate uses of statistics in "the news", and throughout the course.
University of the Pacific
Stockton, CA 95211