How Not to Lie with Statistics: The Good, the Bad, and the Average

The formula is painfully familiar— “According to a recent survey by X from the University of Y,” followed by a statement about married couples’ tendency to get bored with their sex lives (ABC News), or thin women’s tendency to think themselves fat (Psychology Today), or older people’s tendency to become increasingly religious (Harris Poll, 2006)—or any number of statistically proven and so presumably unquestionable claims about human nature.

But what are you really supposed to make of such claims? To take the first case, should you really avoid wedding bells like they toll the death of all fun?

Many of the reputable psychological surveys so often cited by the media seem to think so. The ABC News “sex survey” concludes that “sexual excitement declines with the duration of a marriage or committed relationship—losing the spark.” The evidence for this rather disheartening generalization? “Among couples who’ve been together less than three years, 58 percent call their sex lives very exciting. At more than 10 years, only half as many, 29 percent, say so.”

The statistical interpreters don’t seem to recognize that those other 29 percent are real people, in long-term committed relationships, who report exciting sex lives—so clearly the “losing the spark” phenomenon is not universal or inevitable. Yet the mistake is common; readers see an article’s sweeping conclusion about the decline of sexual excitement with time, and perhaps come to fear (or even avoid) long-term committed relationships—thinking they know where it leads.

This assumption that statistical tendency equals human nature abounds in recent news: Science Daily reports that “men overcompensate when their masculinity is threatened” (all men?); Medical News Today maintains that people’s beliefs about food “can seldom be shaken by rational arguments” (all people?); Associated Content pontificates that the more time someone surfs the web, the more socially awkward he or she will be. But what do any of these claims mean?

To answer this question, we must distinguish between a statistic and a fact of human nature. Most psychological studies simply report such findings as statistical averages, without implying that they are universal. But then something happens: the media pundits who report the findings lose the distinction between statistical average and universal truth. What starts as a conclusion about the way most people happen to be, turns into a prognosis about they way you have to be.

The plain fact is, you don’t have to toe the line of the statistic. People can help their behavior. Plenty of women realize from common sense and observation that they don’t need a Mary-Kate Olsen waist to be attractive. Plenty of web surfing enthusiasts maintain bustling social lives without hiding behind their screens as a defense against confrontation. And as anyone knows who has ever browsed the stacks of a Barnes and Noble, the shelves overflow with advice and innumerable techniques for spicing up one’s sex life – so that any committed couple can enjoy decades of adventurous love-making.

This is not to deny that the statistical tendencies described by these studies do exist. It is true that, for instance, there is real potential for internet junkies to use the web as a crutch, avoiding live interaction in favor of the less stressful, more safely smokescreened medium of virtual communication. But while the internet offers this potentially enticing escape to those who choose to exploit it, the decision is up to every individual user. Cyber communication methods like AIM and Facebook can just as easily be used to supplement and enhance live social interaction as to replace it. Even the shyest among us, who sincerely dread the perils of a live conversation, can choose to confront their fears, and to reap the rewards that eventually follow. There is no cyber-menace stealing their will and compelling them to hide behind their text messages, avoiding the risk of genuine intimacy. Unsurprisingly, however, a statistical majority choose the less risky (albeit less gratifying) route. This is the truth that’s captured by the article’s statistic.

What statistical surveys actually tell us, supposing they are accurate and scientifically valid, is what will happen if we let ourselves be driven by external influences—if we don’t judge the influences for ourselves and choose the best (if not easiest) ways in which to conduct our lives. The theoretical articles and newspaper op-eds that cite them should be conveying the opposite of their current message. Rather than imply that the statistics are inescapable, and then blame the external influences that allegedly determine human behavior, they should treat statistical trends as a reminder of the importance of independent thought: if and when you don’t consciously direct your life, external influences take over. People should be told that other people, on average, tend to absorb cultural standards unthinkingly—and then be taught that they are free to do otherwise.

You are not fated by statistics. You always have the choice of becoming one of the “outliers,” those independent thinkers who don’t just blindly fall where they will on the standard normal curve—but rather take control of the variables governing their destiny.

Gena Gorlin is a senior undergraduate psychology major at Tufts University, currently working as a clinical and research intern at Mass General Hospital.

Add Your Comments
Post Tags
Written by