Abstract Graphic Image

Dear Colleague,

I would like to show you an effective new approach to teaching the introductory statistics course. I have written three papers about the approach, the abstracts of which follow:

The Introductory Statistics Course:
The Entity-Property-Relationship Approach

Abstract: This paper proposes six concepts for discussion at the beginning of an introductory statistics course for students who are not majoring in statistics or mathematics. The concepts are (1) entities, (2) properties of entities, (3) variables, (4) a major goal of empirical research: to predict and control the values of variables, (5) relationships between variables as a key to prediction and control, and (6) statistical techniques for studying relationships between variables as a means to accurate prediction and control. After students have learned the six concepts they learn standard statistical topics in terms of the concepts. It is recommended that each concept be taught in a bottom-up fashion with emphasis on concrete practical examples. It is suggested that the approach gives students a lasting appreciation of the vital role of the field of statistics in empirical research.

This paper contains 31,000 words and 190 references and is available over the Internet in four formats:

 

The Entity-Property-Relationship Approach to Statistics:
An Introduction for Students

Abstract: This paper introduces the "entity-property-relationship" approach to science and statistics at a level suitable for students in an introductory statistics course.

This paper contains 21,000 words (including 24 exercises) and is available over the Internet in Adobe Portable Document Format. (306 kilobytes. For information about the free Adobe PDF reader, click here. Click here if you have a problem viewing the PDF document.)

 

Eight Features of an Ideal Introductory Statistics Course

Abstract: This paper discusses the following features of the author's ideal introductory statistics course: (1) a clear statement of the goals of the course, (2) a careful discussion of the fundamental concept of 'variable', (3) a unification of statistical methods under the concept of a relationship between variables, (4) a characterization of hypothesis testing that is consistent with standard empirical research, (5) the use of practical examples, (6) the right mix of pedagogical techniques: lectures, readings, discussions, exercises, activities, group work, multimedia, (7) a proper choice of computational technology, and (8) a de-emphasis of less important topics such as univariate distributions, probability theory, and the mathematical theory of statistics. The appendices contain (a) recommendations for research to test different approaches to the introductory course and (b) discussion of thought-provoking criticisms of the recommended approach.

This paper contains 18,000 words and is available over the Internet in Adobe Portable Document Format (158 kilobytes. For information about the free Adobe PDF reader, click here. Click here if you have a problem viewing the PDF document.)

 

I welcome your comments about the ideas, either in the e-mail list EdStat-L (= the sci.stat.edu Usenet newsgroup) or via e-mail.

Thank you for your interest.

Don Macnaughton < donmac@matstat.com >

Articles

The following articles were posted to (or announced in) EdStat-L and sci.stat.edu on the indicated dates. The latest articles are at the end.

  • On July 17, 1998 I posted the abstract of the "Eight Features" paper (available above) to sci.stat.edu and EdStat-L. Following are my responses to comments about the abstract and paper:
    • response to comments by Dennis Roberts (July 23, 1998) re teaching univariate distributions at the beginning of the introductory course; are univariate distributions boring? when univariate distributions are necessary
    • response to comments by Mark Myatt (July 30, 1998) re teaching univariate distributions at the beginning of the introductory course; the role of univariate distributions in exploratory data analysis
    • response to comments by Rolf Dalin (July 31, 1998) re omitting teaching univariate distributions at the beginning of the introductory course; other goals and features of the introductory course
    • response to comments by Gary Smith (November 23, 1998) re three interesting examples of univariate distributions (election predictions, average body temperature, and the speed of light); how these examples fit into the rubric of relationships between variables; why questions about univariate distributions are less important in an introductory statistics course; APPENDICES: are relationships "between" or "among" variables? physical constants versus entities, properties, and relationships; the role of constants (physical and otherwise) in empirical research; ways of viewing empirical research projects
    • response to comments by Rossi Hassad (November 25, 1998) re whether there is a necessary logical progression from univariate distributions to relationships between variables; whether the constructivist view of learning implies that we must first discuss univariate distributions; the relationship between relationships between variables and univariate distributions; the simplicity of the issues
    • response to more comments by Dennis Roberts (May 2, 1999) re the breadth of application of statistical ideas; whether almost all empirical research projects can be usefully characterized as studying relationships between variables; some statistical methods that can not be easily characterized as studying relationships between variables; evidence that univariate distributions alienate students; why some students find statistics to be boring; how to make the introductory statistics course fascinating
    • response to comments by Karl L. Wuensch (May 9, 1999) re two examples of univariate distributions that may be of interest at the beginning of the introductory course; the notion of "payoff" as a key to why relationships between variables are more interesting than univariate distributions; why teachers continue to discuss univariate distributions; a situation in which univariate distributions are of interest
    • response to comments by Herman Rubin (May 16, 1999) re the importance of univariate distributions in actual problems; what is needed to discuss relationships between variables? can a relationship exist between random variables? a definition of a relationship between variables; APPENDICES: should it be "relationship" or "relation" between variables? testing for a relationship between variables in Francis Galton's height data
    • response to comments by Jan de Leeuw (October 31, 1999) re the usefulness of the concept of 'external world'; a comparison of two definitions of the role of statistics; confusion about hypothesis testing; alternate formulations of the distinction between univariate distributions and relationships between variables; the importance of univariate summaries; a survey of the types of statistical statements in a newspaper; statistics as a service activity; the distinction between the descriptive and action-prescriptive aspects of statistics
    • response to comments by Bob Hayden (July 23, 2000) a comparison of four popular types of plots to show the relationship between a continuous response variable and a discrete predictor variable; the order of teaching (a) the concept of 'relationship between variables' and (b) data plots in an introductory statistics course; teaching students the underlying conceptual principles of statistics versus teaching them the underlying mathematical principles; a practical first detailed example of a relationship between variables in an introductory statistics course; a system for collecting student course-work-time data; the visual t-test
    • response to comments by Ronan Conroy (February 6, 2001) the substantial contributions of André-Marie Ampère; Ampère's elucidation of the scientific method; a very brief history of the use of the concept of 'relationship between variables'; a possible model of animal learning in terms of relationships between variables; the role of relationships between variables in an approach to machine learning; the concepts of 'population' and 'sample' in empirical research in the physical sciences; why statistical methods are used less often in the physical sciences; how statistical methods might have helped in the cold fusion controversy; the undefined nature of the concepts of 'entity' and 'property'; the concept of a "measure" of the value of a property; the concept of the "true" value of a property; formality and informality in the progress of science; comments on undefined concepts in physics
  • Definition of "Relationship Between Variables" (response to comments by Jan de Leeuw, Herman Rubin, and Robert Frick; January 28, 2002) evaluation of seven general definitions of the concept of 'relationship between variables'; a recommended definition of "relationship between variables" for students who are not majoring in statistics or mathematics; must we force students and clients to clamber from their own "space" into our statistics "space"? must we discuss univariate distributions or underlying mathematics at the beginning of an introductory statistics course for students who are not majoring in statistics or mathematics?
  • The Most Exciting Talk at the 2003 Joint Statistical Meetings (August 22, 2003) comments about Candace Schau's SATS (Survey of Attitudes Toward Statistics); the promise of experimental research in statistics education
  • The Most Exciting Talk at the 2005 Joint Statistical Meetings (August 25, 2005) comments about Mack Shelley's talk "Education Research Meets the Gold Standard: Statistics, Education, And Research Methods after 'No Child Left Behind'" about experimental research in education
  • New Response to Comments from Mike Palij (February 11, 2007) Some basic ideas of science. Novelty of experiments in education. Problematic nature of experiments in education. Necessity of experiments in education. Changing attitudes toward experiments in education. Opportunities for experiments in education. Appendices: A. Eight problems in experiments in education. B. Courses about experiments for education researchers. C. Can human performance or behavior be predicted from a person’s race? D. Specifying a repeated measurements analysis of variance.

Return to top.

Sums of Squares in Unbalanced ANOVA | MatStat