If you torture data sufficiently, it will confess to almost anything.
− Fred Menger
There are lies, damn lies, and statistics.
− Mark Twain (also attributed to Benjamin Disraeli)
Often TV news carries a breathless report: “A new study shows that…” Usually the “study” claims to reveal some new threat. We get invaluable information from scientific research. But what guarantees that the new “study” is, in fact, scientific, and not driven by some agenda?
“Crazy” soldiers and veterans.
Films and TV repeatedly show current or former service members as alcoholic, drug-addicted, unemployed losers who are often suicidal or homicidal. Examples include “In the Valley of Elah,” “The General’s Daughter,” and “The Hunted.”
Not to be outdone in the “bash-the-vets” race, the New York Times published a “study” of 121 homicides in America that are alleged to have been committed by persons who had served in Iraq or Afghanistan. The obvious implication is that veterans are unstable and potentially dangerous. Not explained is why this did not apply to the “greatest generation” that fought the much bloodier World War II.
As John DiIulio pointed out in the Weekly Standard, accused murderers are counted with convicted ones. Also, about 749,932 veterans had been discharged as of the time of the “study.” So 99.98% of veterans of Iraq and Afghanistan have not committed homicides. Why not write about them?
The homicide rate for veterans is about 16.1 per 100,000, while that for non-veterans of similar ages is about 20 per 100,000. The homicide rate for non-veterans is about 25% higher. This is no surprise – people with records of serious crime, mental illness, or drug abuse are ineligible for the military. “The crazy veteran” is a myth perpetuated by anti-military ingrates, who see no problem in slandering the guarantors of their freedom of expression.
Iraqi civilian deaths number “601,027.”
The Lancet, a leading medical journal in Britain, published a “study” claiming to show that at least 601,027 Iraqi civilians have died since the U.S.-led invasion – that is, died by violence, in excess of deaths before the invasion. This greatly exceeds the estimate of 47,668 deaths recorded by reputable people on the ground.
Thus Lancet claimed that over 12 times more civilians died than official sources had stated. Granted, most civilian deaths were caused by terrorists, not by coalition forces. But the obvious point was that the war was disastrous for Iraq. This point went very well with Lancet’s anti-war, anti-American, leftist bias. Previously, Lancet had claimed that America’s homicide rate is the highest of all industrialized nations, which is a blatant untruth.
Now the New England Journal of Medicine, itself liberal and anti-war, publishes a “study” claiming 151,000 excess civilian deaths. Although this is three times higher than the official total, even the New England Journal finds that Lancet’s “study” is a gross overestimate. How many Iraqi civilians have died in the war? I don’t know. But I’ll accept the 47,668 figure, because “scientific” journals have proved themselves to be subject to bias so severe as to produce wildly exaggerated “studies.”
Even worse, how can I believe the articles reporting new drugs and treatments? I know from personal experience that studies presenting positive results for new drugs or treatments often appear in top journals, while studies that present negative results are relegated to second-rate journals, if they are published at all. Objectivity? We don’t need no stinkin’ objectivity. We’re scientists!
The “costs” of smoking.
How are these “costs” calculated? By taking the cost of caring for every smoker with heart disease or cancer of the lung or throat, and adding their lost wages. This assumes that if they hadn’t smoked, these people would either have lived forever, or dropped dead suddenly − and in either case needed no medical care.
A friend of mine was a smoker who died at 57 after a six-month struggle with lung cancer. He died still contributing to Social Security. I miss him. But if he had never smoked, he still would have died, probably of heart disease or cancer. He probably would have lived to 79 and collected Social Security for 14 years. He likely would have had a long-term illness such as a stroke or hip fracture, paid for by Medicare. He probably would have required long-term care.
If my friend had never smoked, he probably would have cost the system more, not less.
Smoking is bad because it causes disease and premature death, not because it costs money. Many economists admit these facts. But what’s more important − facing reality, or pushing an anti-smoking agenda?
Just one more thing: What about marijuana? Since it contains the carcinogens of tobacco, why doesn’t it evoke the same loathing? Why does no one seem worried about its costs – both physical and mental? Why does no one seem worried about the effects of second-hand marijuana smoke, especially on young people? Oh, I forgot – marijuana is “medicine.” Never mind.
Gun control.
The New England Journal published a “study” that reported a lower suicide rate for ages 15-24 in Vancouver than in Seattle, and attributed this to stricter gun laws in Vancouver. The “study” is widely quoted as proving that strict gun laws reduce suicides in young people.
But noted in small print was that the suicide rate for ages 35-44 was higher in Vancouver.
Does anyone claim that strict gun laws increase suicides in older people? Of course not. In fact, the overall suicide rates were similar for both cities, so there was no relation shown between gun laws and suicide. The study was negative, but reported as if it were positive − because it suited the agenda of those who want to ban guns.
Right-to-carry laws.
In 42 states, people without a record of crime or mental illness can get a permit to carry a gun, provided they pass an appropriate course. Dr. John Lott studied all 3054 U.S. counties and found that violent crime decreased where law-abiding citizens could be armed.
Lott’s massive study was reviewed in the New England Journal. A known opponent of gun rights was chosen as reviewer − a breach of editorial objectivity. The reviewer approvingly quoted a “study” that picked a sample of counties and claimed that violent crime rose there after right-to-carry laws were passed.
Statistics teaches us to pick a sample that represents the whole population. This “study” deliberately picked a sample that did not do so. A sample that does not represent the population is invalid. Lott studied all U.S. counties, which clearly yields more definitive data than studying some counties − especially ones picked to “prove” the author’s bias.
But the journal’s editors seemed more interested in their anti-gun agenda than in real research on violence.
Crime statistics.
Data for the early 20th century are incomplete, but they show that the U.S. homicide rate was lower than it is today, even after the recent fall. This is embarrassing. It means that we have to explain why we are killing one another more often now, despite less poverty, less racism, and more gun laws. It means that we have to look at broken families, absent fathers, and declining moral values.
How do we avoid this embarrassment? A scientist “corrected” the data until the homicide rate early in the 1900s is as high as today’s rate. This “corrected” rate is accepted uncritically and quoted widely, while the actual figures are often ignored.
The bottom line.
All these examples twist the data in a liberal direction. Why? I believe the reason is incentive. Most university faculties, editorial boards, and granting agencies are liberal. Researchers want to be promoted, to see their work published, and to get grants. If gun control is popular with editors, submit articles promoting it. If global warming is embraced by granting agencies, make sure your results support it.
In short, money (often our tax dollars) is buying “research.” Why do we howl with outrage if a drug company does this, but we remain silent when our own money is used for the same dubious purpose?
Sometimes actual fraud is involved. Sometimes data are “cherry-picked” to select only what supports the agenda, while negative results are ignored. Sometimes key information is omitted. And sometimes “corrections” are made that push the figures in the “correct” direction.
But in all cases, people are deceived. Real problems are ignored while phony problems occupy our attention. Real solutions are overlooked while false solutions are made to seem effective. Things were bad enough when the liberal agenda invaded the social sciences. But now it is affecting the biological and physical sciences as well, and that is intolerable.
The essence of a scientific hypothesis is that it can be refuted. Einstein declared, “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.” Contrast this with Al Gore’s comment on global warming, “The debate is over.” Which Albert do you choose as your role model: Einstein or Gore?
●
Contact: dstol@prodigy.net. You are welcome to publish or post these articles, provided that you cite the author and website.
www.stolinsky.com