We are confronted with statistics in the news wherever we turn: television, radio, newspapers, magazines and the internet. It can be hard to sort out what
meaning to make of the numbers, especially when there are competing
statistics or interpretations.
Sociology professor Joel Best provides advice on recognizing
suspicious statistics in Stat-Spotting.
This is by no means a technical or mathematical guide to statistics. It is aimed
squarely at the layman who is confronted by statistics in the news and from the
mouths of politicians or experts.
A good place to start is
with a bit of advice that Best puts toward the end of the book (this isn’t
inconvenient; it is a short book). If a number seems shocking, unbelievable or
far outside of what your own experience might lead you to expect, it is
probably worth digging into it some more.
Not every bad statistic
is the result of bad faith. By the time a statistic reaches the public, it has
been through several hands. It starts with some research, which may be
undertaking by a fairly neutral party or by an advocate. In either case, they
have a motivation to get attention for their work. Someone has to bring a study
to the attention of the media, and they may add a layer of interpretation on
the statistics. Finally, reporters, editors and producers are looking for
stories that are sufficiently interesting or important to draw an audience.
This is a process that
can introduce mistakes, even unintentionally, and bring sensational statistics
to the fore. Many of these people don’t know any more about research methods or
statistical analysis that you. The math and logic of statistics, especially when it
relates to probability, can be counterintuitive, and even
professional researchers sometimes don’t have a solid grasp on it. Of course,
some of these people are producing statistics with the intent of supporting a
particular point of view.
Best points out 32 ways
in which the statistics you see may have a problem. These are easy to grasp and
don’t involve much if any math. He presents them in simple terms, and in each
case provides an example from the news.
There are a lot of
demands for our attention and action, and statistics are often cited as part of
these appeals. It is helpful to approach these numbers with some skepticism. Stat-Spotting provides accessible tools
for testing the reasonableness of the statistics we come across day to day.
If you’re interested in
this book, you may also be interested in
Best, Joel. Stat-Spotting: A Field Guide to Identifying
Dubious Statistics. Berkeley, CA: University of California Press, 2008.
No comments:
Post a Comment