Confused by studies? Watch this.

[BendtheBar]

Science For Smart People - YouTube

Science For Smart People - YouTube

[Abaddon]

Watched the whole thing. Loved it.

[BendtheBar]

I learned a lot. It really makes you look at studies that are posted on sites like this through another lens.

[Off Road]

I never listen to studies. For every study that says one thing, there's another study that says the exact opposite. Learn from experience, but I think I said that before.

[Kuytrider]

Originally Posted by Off Road

I never listen to studies. For every study that says one thing, there's another study that says the exact opposite. Learn from experience, but I think I said that before.

Definitely true. As a writer, I research a lot of things and I have written on medical topics a few times (nothing serious!). The amount of contradictory information one finds is startling. Example, I read that high progesterone increased a risk of cancer only to find other 'studies' which said it reduced the risk of cancer!

[MC]

The researcher in me wants to debate, but I know that the work I do and that my colleagues do is better designed than a lot of "studies." So I agree with the vibe of the video and people have to be careful consumers of all information.

The findings of no one study are ever definitive. It takes decades of replication to even get to a point where we can trust results to 999/1000 cases with clinical trials and typically only to 95/100 cases in "action research" (which is not as tightly controlled). In research, we can never "prove" something 100%.

I will say this: don't throw out the baby with the bath water. While not ubiquitous, there are a few good studies on nutrition and exercise.

[BendtheBar]

Originally Posted by Off Road

I never listen to studies. For every study that says one thing, there's another study that says the exact opposite.

I think this video does a good job of explaining why. The example where he features 3 groups of subjects is particularly interesting to my inner-statistician.

[Tannhauser]

Originally Posted by *MC*

The researcher in me wants to debate, but I know that the work I do and that my colleagues do is better designed than a lot of "studies." So I agree with the vibe of the video and people have to be careful consumers of all information.

The findings of no one study are ever definitive. It takes decades of replication to even get to a point where we can trust results to 999/1000 cases with clinical trials and typically only to 95/100 cases in "action research" (which is not as tightly controlled). In research, we can never "prove" something 100%.

I will say this: don't throw out the baby with the bath water. While not ubiquitous, there are a few good studies on nutrition and exercise.

Agree 100%.

MC's second paragraph is the key thing that is missing from many people's understanding of science. That includes the vast majority of journalists. {EDIT: Jeez, that sounds patronising, but my experience is that many don;t really get this at all]

A really, really, really good book on dissecting research studies, and understanding research and science in general, is Ben Goldacre's 'Bad Science.' It's very funny, readable and backed up with genuine intelklectual horsepower.