WASHINGTON -- From a little office on L Street in this unrivaled capital of hype, the Statistical Assessment Service plies the debunking trade.
That scary fact you read someplace? Based on a flawed analysis of the data.
The latest link between a suspected toxin and disease? Not statistically significant.
That intriguing tidbit you heard? An overblown claim from an advocacy group.
Hey, say David W. Murray and his associates, half the stuff "everybody knows" is not supported by the facts. Halloween is not dangerous for children. Suicides do not rise at Christmas. Full moons do not produce a run on emergency rooms or obstetric wards.
Murray, an anthropologist by training and research director for STATS, as it calls itself, scrutinizes the U.S. media for the unproved and the preposterous. He will never run short of targets.
"I've got my pith helmet. This is my village. I study the natives," Murray says. He sweeps a beefy hand at stacks of news clippings with suspect numbers circled in red ink, more evidence of the crimes of innumerate reporters.
In the four years since it was born of another Washington media-criticism group, the Statistical Assessment Service has knocked down many a flawed study and spotlighted many a journalistic howler.
Its annual "Dubious Data Awards" have been reported by the Washington Post and the Los Angeles Times. Some 1,000 journalists and policy-makers get the group's monthly newsletter, a readable collection of gaffes and goofs, and more consult its Web site (www.stats.org).
Reporters call for a guide through statistical thickets.
But the service struggles against a perception that like so many of its targets, STATS, too, has a political agenda. Much of its funding comes from conservative foundations. Many of its statistical critiques raise questions about whether alleged health and environmental hazards are as serious as media reports claim.
Murray, 52, an intense, red-haired, red-bearded man who spent years doing field work on Indian reservations and speaks Navajo, says he has no political ax to grind.
Some of STATS' reports have challenged questionable claims by conservative sources, such as Republican Sen. Jesse Helms of North Carolina and the Weekly Standard, he notes.
"I'm led by science. I believe in the numbers," he insists.
Some of his reports, however, suggest how hard it can be to stand aside as a neutral arbiter of numbers in a polarized debate. By taking skeptical stances on topics from needle-exchange programs to global warming, sometimes in newspaper opinion pieces, STATS has flirted with becoming one more interest group, a danger Murray says he recognizes.
Early in the job, he says, he was invited to an elegant lunch by a well-known Washington lobbyist. Only late in the conversation did it become clear that the lobbyist hoped to persuade STATS to do some work for the tobacco industry. Murray said no.
"As I got out of the cab, the guy said: 'Kid, you're gonna leave a lot of money sitting at the curb,' " Murray recalls.
If STATS goes after environmentalists' overblown claims more often than those of corporations, that's because fewer questionable corporate claims make it into print, he says.
"Journalists are drawn to stories that will get them front-page coverage -- alarm and catastrophe," he says.
The health, environmental and advocacy groups that offer such stories "tend to get somewhat less critical coverage," he says. With companies' claims, "journalists are already wary, as they should be."
The statistical service was created in 1994 by S. Robert Lichter, a political scientist who runs the parent group, the Center for Media and Public Affairs. Lichter remains president of STATS and handles all fund raising.
Lichter acknowledges the rightward tilt of key backers. "The conservative foundations fund you first because they hate the media the most," he says. But he notes that money for the center also has come from the Latino rights group La Raza, the U.S. Commission on Civil Rights and the Ford Foundation.
The idea for STATS, Lichter says, came from his observation of the growing role of numerical data in crucial policy debates. "Journalists are deluged with numbers representing findings in fields they're not familiar with," he says.
There is a natural conflict of roles as the media report science, Lichter says: "If a scientist sees 'man bites dog,' the initial response is, 'This doesn't fit what we know. We have to analyze this further.' The journalist's response is, 'Wow! What a great story!' "
Under Murray, who earned his doctorate at the University of Chicago and has taught at several universities, STATS has evolved into a slightly zany science teacher to the wayward press. Currently a three-person operation with an annual budget of about $450,000, it delights in deconstructing the alleged news.
Sometimes that is not difficult. "Teen Drug Use Dips Down," the Associated Press reported at 2: 57 a.m. one August morning last year. Six hours later, AP updated the story with a new headline: "Drug Use Rising Among Young Adults."
The point was not that many young Americans rose at dawn that day to get high, but that good news or bad news could be plucked from a single, complicated report, STATS said.
In a Centers for Disease Control and Prevention survey of women, the largest number -- 46 percent -- said they believed the biggest risk to their health was breast cancer, the subject of intense media coverage. Only 4 percent named heart disease.
The facts are pretty much the reverse, STATS pointed out: 479,000 women died of heart disease in 1996, edging out all cancers as the No. 1 killer; just 43,000 died of breast cancer.
STATS enjoys shooting down bogus trends. Borrowing from a letter to Nature, STATS undercut a July Newsweek cover story asserting that "a growing number of scientists" were religious believers. Citing a survey of top scientists conducted in 1914, 1933 and this year, the STATS report showed that the proportion of scientists who believed in God or human immortality has steadily declined.
STATS regularly quibbles with the media's alarmism. When detailed AIDS case numbers came out last year, a typical headline was that in the Chicago Tribune: "Women Accounted for 19 Percent of AIDS Cases Their Highest Proportion Yet."
STATS' commentary noted that women's proportion of cases of the acquired immune deficiency syndrome was rising only because the number of cases in men was falling fast. The real headline, STATS said, should have been: "AIDS Deaths Falling, Disease Has Crested in U.S."
Richard J. Gelles, a sociologist at the University of Pennsylvania who studies domestic violence, says: "In my field, David has been scrupulously accurate. He's shooting the messenger because the messenger deserves to be shot."
But Hal Morganstern, an epidemiologist at the University of California at Los Angeles whose study of cancer deaths among California workers exposed to radiation at work was skewered by STATS, says Murray is guilty of the same ignorance and bias of which he accuses the media.
"This guy knows nothing about epidemiology, believe me," Morganstern says. "I realized what's behind it is a very severe political agenda."
Murray replies that his initial dart was aimed at a flawed AP report on Morganstern's study. But when he analyzed Morganstern's radiation study in detail, he says, he found "cherry-picking of data" and overstatement of tiny risks.
"It seems to me a weak case, selectively presented and selectively reported," Murray says.
Odds are the debate won't end there. "Statistics," says an aphorism quoted in a STATS newsletter, "means never having to say you're certain."
Cheer up.
The Statistical Assessment Service says we can stop worrying about a lot of things:
Parents spooked by the possibility that their children could be given poisoned Halloween candy can relax. A California study found three child deaths nationwide since 1958 initially blamed on such poisoning: One, it later turned out, was deliberately poisoned by his father; a second ate his uncle's heroin; and a third died of a seizure.
It is safe to go back into the water. A report in Nature says only seven people have been reported killed by great white sharks in this century. Four times that number of children have been killed since 1990 when TV sets fell on them. Hence, says STATS, "Watching 'Jaws' on TV is more dangerous than swimming in the Pacific."
Not all disabilities are total. A 1996 report from the National Center for Health Statistics received wide coverage for its finding that nearly 100 million Americans have "chronic diseases or disabilities" and their care accounts for three-fourths of U.S. health spending. But the largest single category, as STATS pointed out: the 32 million people with sinusitis and hay fever.
Pub Date: 10/30/98