Pam Belluck explains the hold-ups with the 7-billion-dollar National Children's Study: "Wanted: Volunteers, all pregnant."
“This study is of the magnitude of the accelerator in CERN, or a trip to the moon — a really big science issue,” said Milton Kotelchuck, a professor at the Boston University School of Public Health and a member of the independent panel. “But if you have a flawed beginning, then you’ve got 20 years of working on a flawed study.”
The study locations (105 counties nationwide) were determined by a political process -- they're strategically located in certain congressional districts. So all of the stuff in the article where Congresspeople are complaining about the unexpected price tag, well, put the blame where it belongs. The sampling scheme seems to me like it will create the mother of all confounding variables, unless they're lucky and none of the environmental measures vary among study counties in the same way.
At any rate, the current problem is that they can't recruit enough babies. In at least one county, they've put themselves in the situation of having to recruit more than 30 percent of births! Imagine -- you've got to convince a third of families (somehow, the father's role in this never gets mentioned...) to let the government follow their child for 21 years.
