[quote]MikeTheBear wrote:
[quote]TwinIron wrote:
[quote]MikeTheBear wrote:
That ergo-log is a gold mine. It’s perfect for guys like me who are not professional strength coaches but consider strength training as a bit more than a mere hobby.
[quote]pumped340 wrote:
[quote]Christian Thibaudeau wrote:
For lots of training related study reviews go to: http://www.ergo-log.com/
While I personally prefer pubmed or google scholar to do my own research, it’s cool to have stuff spoon fed to us from time to time![/quote]
That’s pretty cool. How do you feel about the applicability of most of the studies? For instance I’m sure many studies may find a “significant difference” but in the real world it still might not matter. Or on the other hand I’ve seen studies showing “no significant difference between groups” even though there was a difference of a few kg of fat lost. [/quote]
Hope I’m not out of line for answering since this is CT’s thread. When studies refer to something as “significant” or “not significant” they’re talking about statistical significance. Most, if not all, study results are published in terms of averages between two groups. If a difference in the averages is not significant that means it can be explained away as a result of chance rather than the differences in the diet or exercise program they were studying.
In your example, I’d be surprised if a difference of a few kilograms was not significant, although it’s possible if a study lasted several months. Most studies would report something like “Group A lost an average of 5.2 lbs. and Group B lost an average of 5.6 lbs.” In this example, it’s pretty clear that Group B lost more weight in an absolute sense. However, was the difference of 0.4 lbs. significant, i.e., was it the result of the difference in the diet or exercise program, or was it the result of chance? A difference that small probably would be the result of chance, meaning that if the Groups were reversed and Group B had followed Group A’s diet/exercise program they still would have lost 5.6 lbs.
Yes, I am indeed a huge geek.[/quote]
Be careful with assessing results with the subjective use of “small.” A “small” difference between two groups can be extremely statistically significant, while a “large” difference between the two can be statistically insignificant. A difference of 0.000001 lbs of average weight loss between the two groups could be significant if the standard deviation (the square root of the averaged distances from the mean) of both the groups is small enough and the level of significance at which the researchers test is restrictive enough. (Of course, the weight loss of each individual subject in both groups would have had to been pretty much equal to have such a nice and small SD).
/that guy
[/quote]
Your point is well taken and you are correct. As a practical matter, and something you alluded to, having an SD so small as to make a 0.000001 lbs. difference statistically significant would not be something you would ever find in a real study, so my use of the phrase “small difference” implied that any difference would be well within the SD of the sample. A 0.000001 lbs. difference would mean that one guy forgot to clean his belly button lint. :)[/quote]
Hahahaha XD.
Btw, sorry for hijacking your thread, CT.