Statistical Sensationalism part II

One of my favorite blogs (although I don't understand half of the subject matter) had a post yesterday about statistics and their persuasive abilities. I have been thinking about this topic as well recently because of two instances where I have heard misleading information on television. The first was on an episode of Law and Order SVU where a victim had been subjected to genital mutilation as is common many places in Africa and other developing areas. The medical examiner on the show said that even though it is illegal in this country, some immigrants and aliens still perform the act, and "some 200,000 girls could be at risk." Now I have a problem with this statement. While it is most likely true, the way it was stated made light of the very important "could be" so that it sounded more like 'will be.' The result of this is that viewers might be shocked into believing that 200,000 girls would be subjected to this horrible procedure. The second error is that there is no time frame for when they would be potentially at risk - this year? The next ten years? This lack of reference to a time period makes it even more misleading and sensationalizes something that most likely is not very widespread in this country at all. The second statistic that bothers me is in a series of ads for autism (strangely enough it is narrated by Mariska Hargitay, one of the stars of SVU). In the commercial a typicall family situation will be shown - leaving instructions for a babysitter or teaching a kid to throw a ball, and there will be an opening statistic offered such as the odds of the babysitter needing to call 911 or the kids growing up to be a professional ball player. Invariably the odds are very slim that the normal seeming situation will happen. Then she tells us that the odds of a child being diagnosed with autism are 1 in 150 - much more likely - so we should help the foundatin or test our kids or whatever it is the commercial wants us to do (I usually am so ticked by the statistic that I don't listen to the end). So, does anyone see the problem here? I'll give you a minute to think about it. No? Well here's a hint - the key word is "diagnosed." That's right folks, that scary statistic isn't about the likelihood of your kid actually HAVING autism, just that he might be DIAGNOSED with it by some hack who probably graduated at the bottom of their class in medical school and can't tell the difference between Ausperger's syndrome and a video game induced stupor. Now, I could get started on the topic of the over diagnoses of kids today (what's the chance you will be told your kid has ADD - 1 in 3?), but that is a topic for another day. As for this, I'm going to look into the odds that a child mught actually have autism, and keep on throwing the ball to my kids.