Learning Statistics Through Playing Cards
||Thomas R. Knapp
||Sage Publications, Inc., ©1996
Sage Publications, Inc.
2455 Teller Road
Thousand Oaks, CA 91320
$38.00 (c); $16.95 (p)
This short (98 pages) and understandable book is must reading for all
nonstatistically trained professionals in the justice system or in any authority
where decisions are made affecting the lives of other persons. This would
include most judges, lawyers, physicians, social workers, and some psychologists.
Not only are statistics regularly presented in the media, cited to buttress or
support some claim, but are also frequently offered in evidence that is given,
even if only in the form of claims about "most," "typical,"
consistent with," or "classic" observations or patterns.
Such statements are often given weight and credibility when they should not be
attended to at all.
It is shameful that so many who are involved in doing things to other people are so woefully, abysmally ignorant of statistical inference and the laws of probability but still use fallacious statistical information. There is great error introduced
into what should be the most accurate decision-making process we can generate,
the justice system, when professionals reach opinions or conclusions on the
basis of mistaken concepts of probability and what statistics actually mean.
An example is the physician who testifies that a child has been sexually penetrated on the basis of three signs or indicators, each of which is acknowledged by the physician to be nondiagnostic of abuse by itself. However, the claim is that three of them
together make it a certainty. This is wrong. If anybody goes to
prison on the basis of such a claim it is gross injustice. The reason is
that to get the probability of an event by combining non-independent signs you
must multiply the probabilities of each individual sign, not add them up. When a number less than 1 is multiplied by another number less than 1, the product is always less than either of them. Combining unreliable signs does not increase probability but decreases it.
Such testimony should not be allowed as expert testimony. Lawyers who
need to cross examine experts who make such claims can read this book, do the
assignments, and learn how to impeach any expert who makes these kinds of
errors. There are plenty of them out there and in the courtroom.
This book uses a deck of cards and a series of explanations and assignments with the cards to teach
what statistics are about and how to understand them. In the space of, at
most, three hours of actually working the assignments with a deck of cards, any
judge, social worker, or lawyer could acquire at least enough understanding to
avoid the most egregious errors of judgment often made. The basic concepts
of populations, frequency distributions, sampling, central tendency,
probability, and statistical inference are clearly explained and, most
important, actually demonstrated by the assignments using a deck or cards.
This is an ingenious way to illustrate and show what statistics is about.
All professionals who make decisions about others' lives will benefit from this book.
Reviewed by Ralph Underwager, Institute for Psychological Therapies, Northfield,