Feb. 13, 2007 -- On stories ranging from the value of naps to the ministrations of adoptive parents, it has not been a terrific week for the sober reporting of scientific data.
Then again, it rarely is.
The week started on a suitably flat foot for data reporting when the Sunday New York Times published a piece by its ombudsman criticizing the paper for a Jan. 16 story that said that, apparently for the first time, more American women were living without a husband than with one.
Our internal advice at the time was for ABC News to steer clear of it, because it wasn't reflected in our own survey data or in our reading of the Statistical Abstract of the United States.
Lo and behold, the Times now says its analysis counted 15- to 17-year-olds as women. Indeed, almost 90 percent of the women this age not only were spouseless, but were exactly where you might think they'd be -- living at home with their parents.
Today's new bundle of joy is a study on adoptive parents published in the American Sociological Review. The Associated Press report on this study looks darn newsy: "Adoptive parents invest more time and financial resources in their children than biological parents."
What might have been made clearer is that this is so because adoptive families -- of which there were 161 in the sample -- have higher incomes and more education. When the data are controlled for income and education, there is no significant difference between married mom-and-dad adoptive parents and mom-and-dad biological parents. Both do equally better than other family types.
The report's basic finding is that compared with being raised by a biological mother and father, being raised by an adoptive mom and dad "does not unequivocally constitute a disadvantage" in the allocation of resources to young children.
That is not the same as saying it's an advantage over being raised by two biological parents. In fact, the study says that "We find that the two-adoptive-parent family structure is remarkably similar to the two-biological-parent family structure," and that both are better than alternatives.
The AP report also notes in its lead to the story, and discusses at some length, the notion that this report is "challenging arguments that have been used to oppose same-sex marriage and gay adoption." In fact, the study specifies that because there are too few such families to analyze in the data, it is focused only on "married male-female couples who adopt."
You could also have relaxed this week with the nap study, published in the Archives of Internal Medicine. Again per the AP, it "provides the perfect excuse for office slackers, finding that a little midday snooze seems to reduce risks for fatal heart problems, especially among men."
The study indeed finds a correlation between naps and reduced risk of dying from a heart attack. That's among Greeks.
Among Americans, who knows. But more to the point -- if you haven't heard this a million times, you should -- correlation is not causality (as implied by the phrase "seems to reduce"). It simply means there's a statistical relationship between the two, not that one causes the other.
The strongest effect was observed among working men. It is entirely possible that working men who have time to nap have less stressful jobs -- and that it's the lack of stress, not the lack of naps, that reduces their risk. In other words: Does napping reduce stress, or is it simply an indicator of someone who has a less stressful life, e.g., the kind of job in which they can take a snooze?
The study might have moved us toward an answer by rating participants' lives or jobs by their stress level, and using that as a control variable when examining the role of naps. It didn't go there. Therefore, we can't, either.
I spoke today with Dr. Dimitrios Trichopoulos of the Harvard School of Public Health, one of the study's authors. "It would be absolutely premature" to infer causality in the relationship between naps and lower risk of dying from a heart attack, he said. "Causality is a possible interpretation. But we are in no way there. I would like to see another three or four studies before I expressed some confidence in causality."
These reports recall last week's misreporting of the latest figures on autism from the Centers for Disease Control and Prevention in which a 14-state study has been widely and wrongly reported to be nationally representative, and widely and wrongly reported to have produced a sharply higher estimate of autism prevalence than any previous study.
One common point in all these studies is that they appear to have been well-constructed and based on good-quality data. That elevates them above all the manufactured junk data that clamors at our doors and all too often insinuates itself into our news reports. But still the reporting on these studies is not what it could have been.
One take-away is that news organizations -- including this one -- need to sharpen their efforts to report scientific studies accurately. Another is that, as a news consumer, when you see a report on a study that's of particular interest, you might take advantage of the beauty of the Internet -- and click through to the study itself.