The Autism/MMR Fraud | The Incidental Economist
Ms. McCarthy and her followers may have been sincere in wanting to protect children from Autism and it is sad that they have been mislead into doing more harm than good. Even if the study had not been fraudulent, extreme caution would be needed to establish a causal link between Autism and the MMR vaccine. They followed 12 children after being vaccinated. It's hard to make any generalizations from that small of a sample. I could go on and on about problems with this study but there are lots of larger more sound studies that lead many otherwise brilliant people jump to conclusions about cause and effect (especially when it justifies what they already believe).
Merely establishing an association between two variables does not establish a cause and effect relationship between the two. In my post Income and Life Expectancy. What does it Tell Us About US? the video from the documentary The Joy of Stats compellingly shows an association between income adjusted per person and life expectancy over 200 years in 200 countries. It is tempting to infer that there is a cause and effect relationship between wealth and life expectancy. However even here there are many potential third variables such as with what is that wealth used to purchase? My analysis in that post should shed light on some of those complexities. Stephen Colbert has a humorous take on this fallacy by spoofing Glenn Beck's promotion of gold on his program.
|The Colbert Report||Mon - Thurs 11:30pm / 10:30c|
|Yellowline International, Inc.|
The best way to determine cause and effect relationships is in an experiment. Ideally an experiment is a study where every variable is controlled and accounted for by the researcher in a treatment group and a comparison group. Unknown variables are controlled for by randomly assigning patients or subjects to groups.
When it comes to autism and the MMR vaccine, it is neither practical nor ethical to do an experiment to see if there's a causal relationship. Experiments are often done on animals first when feasible but for mental disorders like autism or schizophrenia there is no good animal model. Correlational studies are often the best that can be done. It's hard enough to interpret even when the study is not fraudulent.