In another thread someone mentioned that Autism Speaks was involved in funding that study. Since they have an interest in over-inflating the numbers in order to get more attention & money, I wouldn't be surprised to hear that the study isn't quite what it's being presented as.
Also:
http://www.huffingtonpost.com/2011/05/09/autism-study-south-korea_n_859231.html wrote:
CHICAGO -- A study in South Korea suggests about 1 in 38 children have traits of autism, higher than a previous U.S. estimate of 1 in 100
http://www.huffingtonpost.com/2011/05/09/autism-study-south-korea_n_859231.html wrote:
It's not clear whether the children need special services or not, other experts said.
"I'm sure some of these children probably could benefit from intervention, but I don't think we could make a statement that all would benefit from intervention," said Dr. Marshalyn Yeargin-Allsopp, chief of the Centers for Disease Control and Prevention's developmental disabilities branch.
Since they say "traits," rather than "diagnosable with," it's unclear whether they used DSM-IV/ICD-10 criteria, some standard research criteria for ASD, or just made up their own criteria (maybe based on having any autistic traits at all). If it's questionable whether the kids need any help, then it sounds like they stretched the definition beyond a diagnosable/needs-services level, which makes the results interesting, but misleading.