I was a math major before I changed my major to psychology, which makes it somewhat understandable that I was attracted to the study of statistics and their use (and misuse). That interest and the relative expertise I gained in the scientific method and its use in the social sciences has brought me a lot of joy, actually, as it continues to broaden my understanding of human nature and its impact on everything from perception to what we read, what we buy and ideas about why. It, unfortunately, has brought me a lot of pain and frustration as well, having now watched through the years how often statistics have been used to manipulate the general population to achieve an end that sometimes benefits and sometimes diminishes the lives of the very same people who fall prey to the manipulation in their ignorance.
I’m not sure if it upsets me more when politicians malign groups of their own constituents with meaningless statistics that are irrelevant to any honest attempt at improving their lives or the effect of unscrupulous marketing people whose primary focus is getting people to buy things on false pretenses or the astounding hubris of those who create the unending domino game of ignorance, posting incendiary garbage that no one seems to bother to question. It leaves me in a constant quandary of trying to figure out what’s driving the behavior, since the only way to stop something is to understand what motivates it, what the payback is, as Dr. Phil says. Is it ignorance or narcissism or sheer obliviousness to the connections between the fates of those outside our circles and those of our own fates? Most of the time I end up deciding that it doesn’t really matter—the results will be the same either way. And they won’t be good for anyone except a few unless we begin to focus on what we don’t know instead of what we think we know. The only thing more potentially damaging to our country than not knowing what we don’t know is not knowing THAT we don’t know in the first place.
For example, in a political poll, suppose that it is reported that 70% of women are in favor of or disapprove of X. (Fill in the blanks with your own X.) That’s a fact, based on a single snapshot, a single question asked in a single period from a hopefully random group of women—random because if you ask too many people within a group of people like Republican women or black women or women who’ve had an abortion or women who make over a certain amount of money, the answers cannot be extrapolated beyond that group. The statistics don’t reflect the wisdom or the rightness or the reality of the subject at hand—only opinions. And not only that, opinions derived from as many unique experiences and beliefs as there are women who share the view.
Then there’s the problem of the 30% who, for whatever reasons, don’t agree. And the rather obvious fact, if you think about it, that if you walk down any street and meet a woman you have never seen before, you could no more tell whether she’s in the 70% or the 30% than you can predict the weather next year on your birthday. And though some of you may want to argue the point to further some esoteric agenda, to say otherwise only proves that you fall in that narcissistic group mentioned above—those to whom it has never occurred there exist equally valid opinions other than theirs.
Statistics do not prove cause and effect—never have, never will. Just because 70% of women believe in a certain way does not provide any information about why they believe in that way or what they will do (or not do) because of their belief. Statistics show correlations, relationships between two variables—in this case that there seems to be a strong likelihood that if you are female, you will feel this way about X. Predictive probabilities, but not explanations for anything.
So, what good are statistics then? They’re invaluable in saving time, in benchmarking change, in understanding if the actions we take move us toward achieving goals or not. They help us predict results based on probabilities, and suggest pathways for further examination. If I give you a valid and reliable psychological test, the items for which are included based on statistics, I have clues to how best to help you, clues to how well you will perform in a given job, clues to whether or not you’ll succeed in college. And if I meet you on the street, I have some data—albeit unconscious—that may help me protect myself. If the statistics suggest that most murders occur at night, the fact that some don’t shouldn’t stop me from keeping an eye out when I venture out in the night to go to the grocery store alone. But that doesn’t mean the guy in the parking lot is out to get me, either.
What we get from statistics are only ever clues. No definitive diagnoses. No pigeonholes. No guarantees. No certainties. And thankfully so. If statistics provided end-all answers, there would be nothing further to investigate. If we were robots programmed only to act in a certain way, this would be one boring place—made up of nothing but Stepford know-it-alls. Ignorant, oblivious know-it-alls.
But we aren’t all Stepford know-it-alls, statistically speaking. And those of us who aren’t can repair our faults. But we have to recognize them as faults. And the fault we must repair first is the narcissistic belief that we know the truth about anyone but ourselves, and even that is questionable.
We must repair the ignorance of our ignorance.