As I was mulling over the "interesting study" thread, It seemed to me that the doping issue has really retarded the ability to conduct valuable research involving elite cyclists. The Daubert standards used by the federal courts are actually a pretty good test of whether a particular study or theory is really scientific: (1) whether the theory can be and has been tested; (2) whether the theory has been subjected to peer review and publication; (3) the known or potential rate of error; (4) the existence and maintenance of standards controlling its operation; and (5) whether it has achieved general acceptance in the relevant community. But applying these standards, there is a huge unknown factor that may or may not be enhancing performance apart from the examined variable such as simply adjusting training or diet. Assuming that the most likely population sample for this type of research, elite cyclists, is possibly tainted by the possibility that study participants could be doping, research has to jump down to the next level or serious amateur competitors or recreational cyclists. The problem is that in addition to any control issues you may have with any sample, with amateurs you have any number of other factors related to their real job, family obligations, genetics, diet, motivation to really see whether a particular variable is significant. In other words, you better have a pretty big sample before you start claiming significant results with this type of population. The truth is, is that kind of research any better than the trial and error used by most better coaches? I.E. my cyclists are getting faster and beating their competitors therefore my methods are valid. If so, the research only satisfies the first two prongs and the last three are really in serious question.
For example, if you ran a regression analysis on the entire field of pro elite cyclists to see whether weight training, or large gear training, or whatever, enhanced race performance, how could you even be sure that significance was related to whatever variable is being examined, as opposed to their particular doping regieme. And that doesn't even get into what kind of doping because obviously there are no peer reviewed studies showing what kind of doping regiemes are most effective against each other because that assumes that athletes admit doping and undergo rigorous peer reviewed studies comparing doping regiemes with certain training regiemes.
And how is a small study sample, even under lab conditions, any better than the trial and error used by a good amateur coach with hundreds of athletes?