Archive for January 21st, 2007

Why I’m not a study-citation guy

Sunday, January 21st, 2007

This is a rant.

I’ve touched this subject several times since the late 90’s in articles, forum posts and seminars. But since I see people quoting unrelated or flawed studies on daily basis at various forums, it’s still as relevant as ever.

So, what’s my beef with studies? Well, for starters:

  • Most studies are performed in an unlikely environment, not applicable in the real world
  • Most studies are performed on an exclusive group of people, and the results cannot be transferred on to other groups
  • Most studies has too few participants to get any noteworthy data
  • Most studies are too short to show any significant data
  • The interpretation of studies is multiplex with many potential pitfalls (just take a look at journalists and their sensational conclusions)
  • Studies are constantly contradicting each other, which leads to another big point…

According to Greek epidemiologist John Ioannidis’ mathematical model, a well-designed study (with no professional bias) has an 85% chance of being right, while a poorly designed study with researcher bias has a 17% chance.
After breaking down all the research data he collected between 1990 and 2003 he concluded that 50% of all published research is probably wrong.

Now, this is why I only cite well-respected and well-backed up studies in my articles and why I never jump on the bandwagon when some new contradicting and sensational study is being published.

However, I still skim trough studies and reports on a weekly basis. Most of it is not applicable in the real world, but now and then something interesting shows up. If it’s relevant I’ll do my own field testing and discuss it with other fitness professionals. However, most of the time those studies only validate what we have been doing for years. When it comes to the know-how, the research community is often years behind innovative trainers and coaches.

Now, in our field of training and nutrition – please don’t cite studies or draw your own conclusions from them unless you know exactly how to interpret research data and has less than several years of experience training people. Leave that to the experienced experts in the field.

Post to Twitter