Following the Mislabeled Trail

If you ever go hiking in a state or national forest, you'll notice little triangles at eye level on certain trees. In case you didn't realize it, they are marking the path for you. You can be reseaonably confident that the rangers are keeping the trails in good shape for you and you can trust that they'll get you back to the parking lot.

The rangers would not intentionally steer you wrong.

There is another group of people we encounter every day who also do not intend to lead us astray but nevertheless are. These are the scientists.

I've often commented that I'll eat all the butter I want and not care what "they--the ubiquitous they" tell me is good or bad for me, because as with the whole butter situation, they often change their mind. Years ago we all ate butter with every meal and dessert. Then they told us it was clogging our arteries and we should use margarine instead. Now they tell us the type of fat in margarine is bad, so eat butter instead. So I'll eat butter.

I'm still reading Jorda Ellenberg's How Not to Be Wrong, and he explains why this phenomenon is so:

When scientists create a hypothesis and then test that hypothesis, they are looking for what is called "statistial significance." That is p<0.05. That means that whatever they are looking for happens 1/20 of the time. So if they are looking for a correlation between jelly beans and acne, if 20 subjects eat green jelly beans and 1 of them gets acne, but this doesn't happen with any other color, then scientists might conclude that there is, indeed, a link between green jelly beans and acne. (This is Ellenberg's example but I have simplified it; hopefully I have done it justice.)

This chapter is important to me (the rest has been interesting, but this one actually affects my life) because I deal with health issues and tend to read health blogs/news/magazines. I am not one who tends to pulling my hair out with worry (it falls out well enough on its own) or even change my habits because of what they say (see butter example above).

But for those of you who take to heart what you read in the news: buyer beware! Just because a scientific study says driving with the windows open causes cancer doesn'e mean it does! It means that at least one person in twenty who has cancer also drives with the windows open. (Again, that's an oversimplfication.)

Ellenberg also points out that scientific studies are supposed to be replicable, meanng if you run a study, some other scientist should be able to run the same study and get the same results. This often is not the case. Ellenberg writes, "In a 2012 study, scientists at the California biotech company Amgen set out to replicate some of the most famous experimental results in the biology of cancer, fifty-three studies in all. In their independent trials, they were able to reproduce only six."

Makes you wonder about studies linking vaccines and Autism or gluten and autoimmune disorders.

As I commented earlier, the scientists are not intentionally leading us astray. Part of the problem is that there are so any variables to control for! Another problem, as Ellenberg points out, is the way these studies are published and the criteria th studies must meet. If you want to know more about that, read the book!

My very unscientific conclusion: eat butter and drive with the windows down.