I wanted to get this out there, my apprehension of so called "scientific studies". The big problem that I have is with linked data studies is the lack of association between the empirical evidence and the effects that are being suggested. If I told you that Arizona has the highest rate of lung Cancer, could you conclude that something in Arizona causes that lung cancer? In fact, the opposite is true. Arizona has such high rates of lung cancer because older people migrate there, for the warmer weather, and older people have higher rates of cancer. If a study were to show that children that play a lot of video games have poor social skills, could you conclude which is the cause and which is the effect? These "chicken and the egg" studies bother me, and I try to give negative correlations a chance to be "innocent until proven guilty"
Quick aside: it has been proven that the egg really did come before the chicken; the eggs came from dinosaurs and the chickens came from the eggs
Subscribe to:
Post Comments (Atom)
This is an important point you raise. At the same time, your intelligent objections are quite often raised by the researchers themselves, or by critics of various studies. Scientists do try to find the truth (sometimes; one hopes more often than not). They also have ways of dealing with these "post hoc ergo propter hoc" errors. Sometimes one thing follows another in time without any causal connection. But sometimes one thing follows another and there is a causal connection. Should we throw out all epidemiological studies because "some" are flawed? Or should we, which I think is the better choice, wade through the difficulties and attempt to find the truth?
ReplyDeleteGood point, and of course I have a solution. It is an interesting enough topic that I will dedicate a whole post, or series of posts to it. Thanks again for the feedback.
ReplyDelete