Quote:
Originally Posted by no kudzu
You can find studies which support every view of every argument. Also true of medical studies dealing with meds, tests, etc. Very hard to figure how to react to these studies. i tend to not believe anything anymore...except my own experience.
|
There are some good ways to evaluate studies -- see where they stand on the
levels of evidence, look for methodological flaws, look to see if the studies are sufficiently powered and if the researchers used appropriate statistical tests, look to see if the study was an appropriate design for the research question it was trying to answer.
The really tricky things are that the nuances of evaluating research aren't always obvious or intuitive even to smart, generally literate people. If you don't study something where research evaluation is a big part of your field or end up in a line of work where that's what you do, there are going to be gaps in that knowledge.
And then there's the problem that a lot of studies get reported on based on their abstracts. Which is a problem for lots of reasons. In some circumstances people reporting on studies don't even have access to the whole study (or
don't read them regardless). Add to it the fact that most of the time the conclusions of a good study look like, "Our results generally support/conflict with the previous work of Doe et al. on X, but further research is needed on Y in order to extend these results to Z," which obviously doesn't make for good headlines, and you end up with things like
this.