During my statistics class this afternoon, the professor stopped to make a statement about the difference between a randomized experiment, like with mice you can control and an observational experiment, like reviewing self reported food intakes and some health outcome.
In the first example, we can say that what we did - what we introduced - caused the change in the mice. In the second example, we can not. We may say that the people who reported eating the most red meat were the ones whot had the most heart attacks, for example.
We then might say that there appears to be an association between red meat consumption and heart disease - in the population. This type of statement is usually followed by, "this study does not allow us to make a cause and effect statement, but it does support the need for further research."
You already know these things because that is why I write this blog.
My professor then gave an example referring to the study released in the past week re: red meat and heart disease. He said that he had heard and read a report in the press and then found the manuscript - the real study and the data analysis of it. He was frustrated because the reporters were making causal statements from an observational study.
He did not have to convince me. The irony was that I have the other study that was reported on the same day - the soda consumption and heart disease - in my book bag. I had been reading it before class so that I could talk about the TRUE findings in my blog!
I tracked the soda one down because Walter Willett was one of the scientists involved and any time his name comes up - you hear about it from me.
All this to say - I am completely spent - and need a little more time to finish the article and wrap my mind around it. (Ok - so you know - the study includes diet soda intake and heart disease outcomes - I do drink diet soda < 2 a day; so I have a personal connection)
More to come.