http://www.epa.gov/air/airpolldata.html http://www.epa.gov/air/airpolldata.html There are probably more APIs out there. Autism numbers by location should be k

http://www.epa.gov/air/airpolldata.html http://www.epa.gov/air/airpolldata.html There are probably more APIs out there. Autism numbers by location should be known. Simple correlation could go a long way to check the consistency of the hypothesis. EDIT: The autism data needed would be age of diagnosis, to map it to the proper air quality metric.