Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.



ActivityBased Statistics
Posted:
Feb 13, 1996 1:14 PM


AP Staters
Since numerous references have been made to the ActivityBased Statistics project, I though it appropriate to share some information with you. This work is a set of approximately 50 activities that can be used to illustrate basic statistical concepts. Most are not computer dependent and can be used in a variety of settings, in class or out, small or large groups. A longer overview follows, if you care to read more.
SpringerVerlag is to publish the work later this spring. Details on publication can be obtained from:
Liesl Gibson SpringerVerlag New York 175 Fifth Avenue New York, NY 100107856
1800springer
OVERVIEW of ABS
Their fastpaced world of action movies, rapidfire TV commercials and video games does not prepare today's students to sit and absorb a lecture, especially on a supposedly dull subject like statistics. To capture the interest of these students, teaching must move away from lectureandlisten to innovative activities that engage students in the learning process. The goal of the ActivityBased Statistics Project (ABSP) is to develop a set of such activities that cover the statistical concepts that are essential to any introductory course. These activities can be used in a variety of class settings to allow students to "discover" concepts of statistics by working through a set of "laboratory" exercises. Whether the "laboratory" is the classroom, the student's usual place of study, or a more formal statistics laboratory, the traditional lectures in introductory statistics should be supplemented or supplanted by a program that requires the active participation of the students, working individually or in groups. Statistics, then, should be taught more as an experimental science and less as traditional mathematics. The activities are organized around the major topics covered in most introductory courses. The overarching topic is Exploring Data. Statistical ideas begin with data, but data should be collected for a purpose. Relating data collection and analysis to the solving of a real problem, much as is done in statistical process improvement, allows exploratory techniques to be used extensively, but with a purpose in mind. Random Behavior is fundamental to the decisionmaking process of statistics, but is a difficult topic for students to grasp. Thus, a number of activities concentrate on developing an understanding of randomness, without going into the mathematical formalities of classical probability distributions. Simulation is key to this process. Related material on Sampling Distributions is included, to be introduced at the discretion of the instructor. Since the amount of required probability is highly variable from course to course, the activities throughout the ABSP are designed to be completed with a minimum of probability. Sampling, such as in the ubiquitous opinion poll, is the most common applications of planned data collection in the "real" world and in the classroom. This topic allows the early introduction of the idea of random sampling and its implications for statistical inference. Thus, it serves as an excellent bridge between probability and inference The difficulty and importance of collecting data that fairly represents a population are emphasized. Estimation and Hypothesis Testing frame the basic approach to inference in most courses, and numerous activities deal with the conceptual understanding of the reasoning process used here. Again, simulation plays a key role as concepts of sampling error, confidence interval, and pvalue are introduced without appeal to formulas. Experiments are the second major technique for planned data collection. The point that experiments are designed to compare "treatments" rather than to estimate population parameters is emphasized by appealing to activities that focus on optimization of factors. Modeling the relationship between two variables, especially through the use of least squares regression, is widely used throughout statistics and should be part of an introductory course. Here, technology is required in order to use time efficiently. Correlation, a topic much confused by students, is singled out for special study. There are far more activities than can be used efficiently in one course. It is hoped that instructors will choose one or two activities from each of the topics so that students are exposed to a range of concepts through handson learning. Each activity has student pages and notes for the instructor. This allows the flexibility to use the activities as class demonstration, group work in class, or takehome assignments. In whatever setting they may be used, the activities should actively engage the students so that they become true participants in the teachinglearning process.



