[quote]Wyzz Kydd wrote:
I am assuming that you used some type of multivariate regression analysis for your study? What was your margin of error and your confidence level?[/quote]
Thanks for the kind words.
I’ve done several regressions with this data over the years. The original (and perhaps still most accurate) was a multivariate exponential regression. In fact, the linear equations presented in the article were actually derived from that original model and then matched with the actual data afterwards. And aspects of that original model were drawn from very large population bases. For instance, from U.S. Army stats of thousands of men it can be determined that men’s weight increase by 3.125% for each inch of height. I compared this to top drug-free bodybuilders and found that to be true as well (with deviations across the board due to bone structure - which can also be analyzed similarly - I used several anthropometric studies to do that).
Then, to help set an upper limit on lean body mass I referenced Kouri and Pope’s work with the fat-free mass index.
Based on only the bodybuilders used in the data set the margin of error would be 6.1% with a confidence interval of 90%. So 90% of bodybuilders will not exceed these predictions by more than 6.1% …but those champion bodybuilders themselves were chosen from hundreds of competitors and were judged to have the best-developed physiques. For instance, Reg Park soundly defeated 42 other competitors at the 1951 Mr. Universe contest. He had, clearly, the most fully developed muscular physique of the 43 competitors on-stage that night (photos of the lineup clearly show this). Each one of the 25 world and national champions used in the “study” can also make that claim. With few exceptions, the winner has the biggest, most balanced muscular physique. If it is assumed that each champion also defeated 42 others when they won their titles then the margin of error would be 3.1% with a confidence interval of 95%. But those 42 bodybuilders were also good enough to make it to national and world-level competition themselves. How many did they each defeat to get there?
So I really can’t put an error margin and confidence level on this that would be applicable to the general training public.
I find it amusing how many people are getting personally angered by their perception that I’m claiming to know their absolute limits of potential. I never made such claim and that is stated clearly in the article. However, what I am clearly saying is that if you intend on surpassing the predictions of the equations you are also intending to be equal to or greater than, with regards to muscular size and balance, the world’s greatest drug-free bodybuilders. Perhaps next year’s WNBF Overall champion is reading this, but 99.9% of the population are nowhere near his level of development (and never will be) and need appropriate goals and expectations from drug-free bodybuilding.
