Saturday, October 22, 2005
CASA built on falsehood, not fact
Family dinner vs. drug use just the latest howler from Joe Califano's fanciful 'U.S. research centre'
You might have seen the former U.S. first lady's anti-drug commercials on American television recently. But it wasn't Nancy Reagan telling kids to "Just say no," it was Barbara Bush encouraging families to just say yes -- to family dinners, that is.
The commercials accompanied the release of a survey by the U.S. National Center on Addiction and Substance Abuse at Columbia University (CASA) titled The Importance of Family Dinners II. The survey, widely reported in newspapers across North America, including The Vancouver Sun, found that the frequency of family dinners is a "powerful indicator" of whether children will use drugs:
Compared to kids who eat five or more family dinners a week, children who dine with their parents twice or less a week are three times likelier to try marijuana, two and half times likelier to smoke cigarettes and more than one and a half times likelier to drink alcohol.
These results led CASA chairman and president Joseph A. Califano Jr. to conclude: "If I could wave a magic wand to make a dent in the substance abuse problem, I would make sure that every child in America had dinner with his or her parents at least five times a week."
Alas, there is no magic wand, but Califano's comments certainly resemble magic. Black magic.
To begin with, one needn't have studied research design or statistical inference to see that the conclusions Califano draws are invalid. It might well be the case that kids who eat with their families are less likely to use drugs, but that is evidence of correlation, not causation.
CASA's report, which is a survey and not a study (as some media have erroneously labeled it), can't tell us anything about causation: We don't know if eating family dinner actually causes children to just say no to drugs, or whether there is a third variable, such as the overall level of family functioning, that influences both the odds of the family dining together and the odds of the children using drugs.
If this third variable explanation is true -- and it is likely -- then sitting a dysfunctional family together at that table won't do anything except perhaps produce a Jerry Springer-style food fight.
Those who are schooled in research design have discovered that the CASA survey suffers from other serious methodological shortcomings. The Wall Street Journal's Carl Bialik, one of the few journalists who decided not to suspend his critical faculties when reviewing the survey, notes that the report had a low response rate: While the researchers began with a pool of 37,000 potential subjects, they only managed to complete 1,000 interviews, and thus their sample is unlikely to be representative of the U.S. population.
Further, and more damning, Bialik cautions that the survey doesn't control for age: Not surprisingly, older kids are much less likely to eat with their parents and much more likely to have tried drugs. Thus age is another confounding variable that influences both eating habits and drug use.
But CASA doesn't see it this way; instead, the researchers conclude that the higher level of drug use among older teens makes it even more important that they eat with their families!
CASA has therefore assumed what needs to proven -- that family dinners, rather than other factors such as age, influence the likelihood of drug use. If Califano really wants to stop kids from using drugs, he should instruct parents not to let their children get any older. Magic wand, anyone?
If you're surprised to learn that a "national centre" on drug abuse -- and one that's "affiliated" with a prestigious Ivy League university like Columbia -- would produce such claptrap, then you obviously don't know much about CASA or Joe Califano.
Califano, a stalwart of the U.S. Democratic party, was Jimmy Carter's secretary of health, education and welfare until he was fired in 1979. In 1992, he founded CASA on the urging of officials at the U.S. Office of National Drug Control Policy, otherwise known as the drug czar.
Califano proved a master at fundraising, garnering major grants from the drug czar and the Robert Wood Johnson Foundation, a latter-day temperance organization. But he received not a cent from Columbia, the institution whose name remains in the title of Califano's group.
According to David Hansen, a world-renowned expert on the sociology of alcohol use and a former alcohol consultant to the Canadian government, Califano took Columbia's snub in stride: By not having any formal academic affiliation, Califano was delighted that CASA could focus on activism rather than academic work.
And despite the fact that many journalists still think CASA is affiliated with Columbia -- this newspaper's report referred to CASA's survey as a "study from Columbia University" -- CASA has clearly eschewed science in favour of activism.
CASA rarely submits its surveys to academic journals, thereby avoiding peer review, which would almost certainly result in the rejection of its work. Instead, CASA "publishes" its surveys on its website and then sends breathless news releases to media outlets, which all too often gobble them up.
Hansen, who has acted as a one-man wrecking crew, discrediting CASA research almost as soon as it's released, provides a few examples of the organization's disastrously bad "science:"
In 2002, CASA reported to much fanfare that underage drinkers account for 25 per cent of alcohol consumed in the U.S despite comprising only 13 per cent of the U.S. population. The U.S. Substance Abuse and Mental Health Services Administration, whose research CASA used to arrive at the 25 per cent figure, hastily confirmed that CASA had more than doubled the true figure, which was 11.4 per cent.
CASA stood by its report, saying that it believed the number must be much higher than 11.4 per cent -- once again assuming what needs to be proven -- and Califano, evidently aware that the best defence is a good offence, issued a news release stating that the proper figure is actually closer to 30 per cent!
In an attack on welfare moms, CASA reported that 27 per cent of women on social assistance are substance abusers. The U.S. Department of Health and Human Services almost immediately chastised CASA for its "seriously flawed" work, and noted that only 4.5 per cent of women on welfare abuse drugs or alcohol.
Once again, CASA was unrepentant, and refused to retract the slur it made against these women.
Perhaps having seen too many Girls Gone Wild videos, CASA reported that 60 per cent of college women who contracted STDs were under the influence of alcohol during sex, and that alcohol is involved in 90 per cent of all campus rapes.
Hansen notes that investigative reporter Kathy McNamara-Meis searched in vain for evidence to substantiate these numbers, and she eventually concluded that they were "pulled from thin air."
Now that's magic. And need I say it? CASA remained un-contrite and never corrected its "research."
There are many other examples of CASA's shoddy work, but it would take a book-length column to document them. Suffice it to say that just because a group calls itself a "national centre," and just because it appears to have an affiliation with a prestigious university, doesn't mean its work is worth reading or reporting. Indeed, most of CASA's work isn't worth the website it's posted on.
That's a good thing, of course, since it means the sky isn't falling as CASA has so often declared. It also means that your kids probably won't become drug dealers or boozehounds just because everyone doesn't make it home for dinner five times a week.
As for me, I think that having dinner at Califano's CASA five times a week would drive me to drink.