The Search for 100 Million Missing Women - An economics detective story. By Stephen J. Dubner and Steven D. Levitt
This is an interesting and well-written article, but it manages to conflate doing regression on large data sets with economics, even though those methods were neither developed by nor are limited to economists. It also manages to bash developmental psychologists for dealing with, in some cases, extremely small data sets.
The problems of drawing inferences based on small data sets are real and well known. The problem that large data sets are often of extremely poor quality is one that is also well known but not, as far as can be judged from this article, to the authors.