The Not Quite Adventures of a Professional Archaeologist and Aspiring Curmudgeon

Wednesday, March 14, 2012

Truths, Damn Truths, and Statistics!

It happened again this week.  While arguing about the safety of a medication, I pointed to statistical data regarding the medicine in question's effecacioussness and rates of pronounced side-effects.  The person with whom I was arguing looked at me and said: "you, Mr. Armstrong, need to remember that each of those numbers you're quoting represents a person!  A fact that you seem too ready to forget!"

No, I don't forget that each of those numbers represents a person.  In fact, if I thought of them as just numbers and not people, I wouldn't be bothering to argue.  I am very well aware that each set of numbers represents real, living people.  People living complex lives in a world full of confounding circumstances that influence their behaviors, their beliefs, and their actions.  Each number or set of numbers represents an individual, unique an irreplaceable in numerous ways, and describes in some way their experience with a medication, controlled substance, interaction with a government agency, etc. etc. etc.

I am aware that these numbers represent people.  I am fully, sometimes painfully, aware of this fact.  It is the people who accuse me of forgetting it who don't seem to grasp this fact.

How do I know this?  Simple: the people who accuse me, and others in my position, of forgetting that the statistics represent people are all too ready to throw those numbers, each of which represents the experiences of an individual, away when they are inconvenient.  When the numbers agree with what a person wants to believe, they are usually only too ready to accept and quote them.  When the numbers disagree, well, they are to be forgotten, or better yet, disparaged.  In effect, the experiences of individuals, of real people, that don't conform to the preferred outcome are ignored, insulted, or hand-waved away.  Invariably, the statement isn't rooted in a desire to see individual cases as having value, but in the desire for the beliefs of the person making the accusation to take precedence over hard facts.

Of course, there are issues that aren't answered with numbers.  There are questions that are not empirical in nature.  And certainly, there are those who ignore this and try to apply empiricism where it doesn't belong.  But these people are few in number, and invariably on the fringe.  Likewise, there are those who will mis-use statistical tools to create false equivalences or show dubious correlations, or simply drown their opponent in a sea of data regardless of how relevant the data actually is (but, of course, these people can be stopped by the correct application of math, not the dismissal of it).  Far more common are those who want to pretend that empirical questions are not empirical, who want to pretend that their personal assumptions trump physical realities.  And you will find these people both advocating and opposing various positions.  You will find them on the political right and the political left.  You will find them in every issue, not matter how well settled the data actually is.

It doesn't matter what issue is being discussed: the safety or effectiveness of medicines, the effectiveness of crime-reduction techniques and measures, the effects of one sex-ed program vs.another as regards rates of STDs and pregnancies, the effects of various substances on pregnancy and on young children, the influence (or lack thereof) of religious belief on lawful behavior, the influence of concealed firearms on violent crime, and so on and so forth.  If it's a question that is best answered with quantitative data based on the experiences of a large number of people, you will find people who readily dismiss the statistics, unequivocally the best way to gather such information and see the patterns.  And those who want to dismiss the statistics will almost always say something to the effect of "you, Mr. Empirical-Data, are forgetting that these statistics represent people!"

But, of course, it's the person who wants the data dismissed or downplayed who truly is forgetting that the data represents people.  It's the person who wants to engage in warm-fuzzy talk about how "we are all individuals, not numbers" who is ready to throw away the experiences of huge numbers of real individuals in order to win an argument.  It's the person that tries to claim that statistics are somehow de-humanizing who is only too ready to dismiss the experiences of their fellow humans in order to avoid losing the argument or ceding the point. 

No, I don't forget that the numbers represent people.  But if you go about proclaiming that those with whom you disagree are ignoring the individual in favor of the numbers, then likely it is you who is ignoring that the numbers represent people.


Allison said...

It's frustrating when people dismiss data like that. There's a guy in my History of American Immigration class that says UCLA is a "liberal think-tank" so he any studies done there are invalid.

Anthroslug said...

Yeah, that's the "Shield from Reality" ( It's no different than evangelical Christians who claim that any information that doesn't comport with their views is "of the Devil" and therefore not to be trusted. It's a lazy person's way of cutting themselves off from having to deal with facts.