MathEd Forum

October 31, 2014

T

THE FIELDS INSTITUTE FOR RESEARCH IN MATHEMATICAL SCIENCES


FIELDS MATHED FORUM MEETING AGENDA

THEME:
Has Chance Been Tamed?

March 26, 2011, 10AM – 2PM
Fields Institute, 222 College Street, Toronto

MORNING PROGRAM:
10:00 a.m. - 10:10 a.m.
Reports: OAME, OMCA, OCMA, CMS, and other.

10:10 a.m. - 11:10 a.m.
Sandy Zabell (Northwestern University, Chicago, IL): Twenty Years After

SHORT DESCRIPTION: In the ancient world chance was often contrasted with the absence of cause (Polybius) or Fate's unalterable necessity (Tacitus). Despite the widespread use then of randomization (and somewhat surprisingly for a civilization capable of producing Ptolemy's Almagest), there was no calculus of chances. After the 1654 Fermat-Pascal correspondence, mathematical probability finally emerged, but at first only in the restrictive setting of equally likely cases.
In the early 19th century Laplace gave chance a purely subjective view of its nature: the universe unfolds in a purely deterministic way; probability only expresses only our ignorance of the underlying causes. Over the next century, however, this view became widely challenged. There were many reasons for this: the discovery of statistical regularities in social data, the rise of Mendelian genetics, the development of quantum physics.
Ian Hacking's 1990 book "The Taming of Chance" gave a fascinating and provocative analysis of the origins of this conceptual revolution, and sought to explain its distinctive character. In this talk, after first briefly reviewing the prehistory of the probabilistic revolution, I go on to critically examine the theses advanced by Hacking in his book, and discuss their subsequent reception (or lack of it).

SHORT BIO: Sandy Zabell is a Professor of Mathematics and Statistics at Northwestern University. His book "Symmetry and Discontents" (Cambridge University Press, 2005) is a collection of a number of his papers on the history and philosophical foundations of probability and statistics. His applied interests include the DNA identification, forensic science, and the legal applications of statistics.

11:10 a.m. - 11:40 a.m.
Georges Monette
(York University): Causality for Good Citizens

SHORT DESCRIPTION:
Samuel Wilks (1950): "Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write," paraphrasing H. G. Wells (1903).

Surely that day has come! But what statistical thoughts are needed for efficient citizenship? One could argue that uncertainty about causality is at the root of most current social, political and scientific controversies. Does human activity cause climate change? How will changes in fiscal policy affect economic recovery? Does using cell phones increase the risk of cancer? In almost every case, controversy subsists because causal information is needed from data that are largely observational as opposed to experimental. Students of statistics learn the dictum "correlation is not causation," a truism that doesn't help them develop the finer judgment necessary to evaluate imperfect evidence from observational data that is often all that is available to inform crucial and urgent issues.
What more can statistics say? We would like to explore whether a good citizen's version of causality can be made simple enough to teach in high schools yet no so simple that it merely breeds belief in new fallacies.
We distinguish between causal versus predictive inference and between observational versus experimental data. Experimental data is ideally suited for causal inference and observational data for predictive inference. For causal questions from observational data, we try to find an approach that leads to a balance between uncritical gullibility and blind skepticism.

SHORT BIO: Georges Monette is an Associate Professor in the Department of Mathematics and Statistics at York University. He obtained his Ph.D. at the University of Toronto in statistical inference. He has had a long association with York's Statistical Consulting Service and has worked on applications of statistics in a wide range of disciplines. Consulting is a form of teaching in which complex concepts need to be conveyed to collaborators and clients who often have little statistical and mathematical background. In addition to teaching academic courses, he teaches workshops on statistical visualization and longitudinal data analysis.

11:40 a.m. - 12:10 p.m.
Alison Gibbs (University of Toronto): Informal Statistical Inference

SHORT DESCRIPTION: Continuing with the theme of the importance of statistical thinking for efficient citizenship, we appeal for increased emphasis in our mathematics curricula on randomness, risk, and reasoning rationally in the presence of uncertainty. In particular, we will consider ideas for expanding the study of statistics in schools beyond graphical and numerical summaries of data and we will show some important recent work about how we can support sophisticated inferential reasoning in very young students.

SHORT BIO: Alison Gibbs is a teaching-stream faculty member in the Department of Statistics at the University of Toronto. Before beginning her graduate work, she taught secondary school mathematics in Ontario. After completing her Ph.D., she held post-doctoral and Assistant Professor positions at York University. In her current position she has taught a variety of probability and statistics courses, ranging from a first-year seminar in statistical literacy to a graduate course in statistical consulting. Her responsibilities also include overseeing and advising the department's Statistical Consulting Service. Her research interests are in applications of statistics, particularly in nutrition studies, and in the development of the full range of skills required of professional statisticians. She is currently the chair of the Statistical Education Committee of the Statistical Society of Canada.

The panellists will describe their experiences from different levels of schooling in Pakistan, Sri Lanka and Czech Republic.

12:10 p.m. -1:00 p.m. LUNCH BREAK
(Light refreshments provided)

1:00 p.m. 2:00 p.m.
Discussion

 


Back to MathEd Forum Page