Monday, August 29, 2011
Critical Thinking Requires a Little Understanding of Statistics
I made a comment a couple of weeks ago about people spending more to buy organic or all-natural products at the grocery store (and other places) without being able to justify the added cost with some tangible benefits. Of course one of the benefits is the good feeling that they are taking better care of their families, but if the evidence shows that there is no difference, it’s money wasted, money that could be used in other ways to make the family’s life better.
What is evidence? Is it a news item on TV or the Internet citing a study that showed health benefits? Listen closely and we find that many things that were bad for you a couple of years ago are now good for you (e.g. coffee, wine and chocolate) and some of the things that were good for you are now either bad for you or have no effect whatsoever – several supplements fall into this category and within the past month a doctor warned that it could be dangerous to be drinking a lot of water when you are not thirsty. (This came with a shocking headline, “Can Water Be Bad For You?”) So the statements and studies come and go. Some evidence is trustworthy and some is not. Just because someone wrinkles his nose and sneers at the limitations of “Western Medicine,” it doesn’t mean that everything the Chinese did for thousands of years is better, or even effective. We can’t necessarily take for truth what we hear or read without at least some preliminary investigation. And we shouldn’t follow the example of the boxer in Paul Simon’s song – “A man hears what he wants to hear and disregards the rest.”
There is a lot to consider when we hear reports. Sometimes the author of a study wants to beat others in order to get credit for a discovery by issuing a press release with “preliminary result.” If these results later don’t hold up we never hear about it. Studies often contradict each other, because they are designed differently or some are poorly designed or those tested differ in some way, or the effect is borderline and the results can go either way. Some studies are sponsored (paid for) by organizations that favor a particular outcome. Is the sample large enough or chosen properly? In health-related studies, it’s a long way from lab rats to people and most conclusions cannot be applied directly.
Do interest groups or news agencies give us enough of this information to make an informed decision? Usually not, and if they do it's at a point where most people have stopped reading. Sometimes the headlines are slanted in a way to get our attention or sway our beliefs.
So it’s left up to us to try to understand the science and statistics – a tall order. Here is an article I found recently to help work through some of the issues. Still, remember when faced with claims of the miracle cure or the cleaner, safer, healthier product, it’s better to be a little skeptical.