we Americans are a statistics-happy bunch. There is hardly anything we
talk about or write about that doesn't have statistics as an integral part
of it — sports, politics, economics, and so on.
Which is OK, I suppose. After all, what harm can statistics do? They're
just numbers.
Well, they can do plenty of harm, when they're used to give an appearance
of solidity to something that is nothing more than pure wind.
By the way, that beautiful phrase about solidity and wind is George
Orwell's creation, not mine. I wish it were.
Anyway, let me give you an example.
Following is a paragraph taken from an article that appeared in San
Francisco Chronicle.
The typical black student with combined Scholastic Aptitude Test scores
between 701 and 850 had a 56% likelihood of graduating from a college where
the average student's score was 900. But his or her likelihood of graduating
fell to 39% at a more selective school, where the average student's score
was 1000.
Now just what did the paragraph say that was useful? In my opinion, nothing.
And yet the writer of the article built an entire thesis supporting so-called
affirmative action around it and other similar paragraphs.
The statistics themselves, assuming they were properly and honestly
gathered, were objective and unbiased. The thesis based upon an interpretation
of those same statistics was not.
Nor could it be, for the simple reason that while a measurement can
be neutral, an opinion, inference, or judgment about the meaning of that
measurement cannot.
Think about it. |
