" Where is the statistical analysis?"
Doesn’t reporting the R^2 value and error estimates take care of this?
I think Dan Empfield summed it up when he said, “You can’t test for this.”
I believe he was talking about aerobars then. I think it applies to a lot of things.
You can test for it but not in this fashion and it may not be doable in the real world because the difference is probably so small as to be “non-detectable” in a statistically significant manner. The reason for the comment is many here and there will read this and think it means something in regards to buying one bike over another. It don’t. If I had written this he would be all over my ass.
“I’d buy him!”
That’s a pretty good idea for a slowtwitch fundraiser. Auction off Andrew Coggan’s services (someone would have to direct him here), the winner gets their training program extensively critiqued, AND gets ringside seats for the next no-holds-barred PC near-deathmatch. All proceeds to charity, of course.
Ken
Andy, I’m sorry Dr. Coggan, is god-like in the sense that He is among us, LOL…
Frank,
Have you ever published the work that lead you to make the following claims? **Get more athletic performance benefit from the same amount of training time. ** The typical cyclist/triathlete can increase power on the bicycle 40% in about 6 months (that is a 7 minute improvement for a 60 minute TT effort!) and the typical runner can **improve marathon pace 1 minute per mile **in less than 3 months! Further, the second year improvement can be just as dramatic,Training for the serious athlete is forever changed!!!
I don’t recall reading how the above claims can be supported.
"What are you talking about? "
That’s funny, becuase I wanted to know the same thing when you posted on that P3 thread regarding aluminum and flex.
That would have been a great thread for andy coggan to chime in on, seeing has how he is such an exerienced hooker rider.
http://www.powercranks.com/newsite/3-other2ndlevelpages/sciencestudies.html
Am I missing something, or are these two studies the same study?
They are the same study. The person who put that together didn’t realize it.
" Where is the statistical analysis?"
Doesn’t reporting the R^2 value and error estimates take care of this?
No
.
Frank,
Have you ever published the work that lead you to make the following claims? **Get more athletic performance benefit from the same amount of training time. ** The typical cyclist/triathlete can increase power on the bicycle 40% in about 6 months (that is a 7 minute improvement for a 60 minute TT effort!) and the typical runner can **improve marathon pace 1 minute per mile **in less than 3 months! Further, the second year improvement can be just as dramatic,Training for the serious athlete is forever changed!!!
I don’t recall reading how the above claims can be supported.
The data upon which these claims are made was gathered by myself and have never been published. Many people at this site have published anecdotal reports and others have reported results directly to me that exceed my claims. I stand by the claims as being typical, if the user uses them as we suggest.
If ever anyone does a long-term study of the benefits of the cranks when used as we suggest then I will probably use that data. If the numbers are better than I now use would that bother you?
Frank
What an awful “study” coming from the same guy who complains about the scientific validity of the PowerCranks study and those studies of one reported here by users. Where is the statistical analysis? Oh, it might be hard to do with a sample of 1. Where is the double blind? How did he control for placebo effect? What on earth does he mean that there might be a “functional” difference but not a statistical difference. To a scientist, if there is not a statistical difference it should be concluded there is no difference.
The data upon which these claims are made was gathered by myself and have never been published. Many people at this site have published anecdotal reports and others have reported results directly to me that exceed my claims. I stand by the claims as being typical, if the user uses them as we suggest.
So let me get this straight…
You rip Coggan when he spells out every little detail of his analysis, yet make these wild-ass claims regarding the benefits of PC’s based on some anecdotal claims gathered by you, not published, and expect us to just take your word on it.
If ever anyone does a long-term study of the benefits of the cranks when used as we suggest then I will probably use that data. If the numbers are better than I now use would that bother you?
…probably???
" Where is the statistical analysis?"
Doesn’t reporting the R^2 value and error estimates take care of this?
No
That was helpful. Can someone explain why not.
So, Frank…
Just curious, but do you go out of your way to rip on everything every well repected expert in the field of bicycle science writes, or is it just anybody who questions the validity of PCs?
Your comments regarding this article of Coggan’s are patently absurd given the nature of the article, and especially so when his summation essentially states that he would not hang his scientific hat on any of it, but that he found it to be personally useful.
Given that he posted this report on Craig’s site, on the forum, which is essentially a group o’ folks chatting about interesting aspects of Bicycle science, the author’s intent certainly doesn’t appear to be to represent this as being in any way authoritative, or anything other than anecdotal.
Too bad Coggan is so freaking methodical that even his BS, just for fun, off the cuff, “interesting note to the guys” stuff reads as being equal science to some of the snake oil being pedalled out there.
I’m not, by the way, accusing you of pedalling snake oil - I don’t have an iron in that particular fire. I am saying that you are carrying around a gigantic axe, and are looking for a place to grind.
Do yourself a favor, and lay off the out-of-context attacks. They are beneath you, and resorting to them cheapens both you and your product.
MH
Tough crowd, eh? The R2 value refers to how well collected data (wattage vs. speed) match to the curvilinear function. Doesn’t matter for comparison of the two frames.
Look at the table w/ the averages: Basically, the averages of the two frames are not far enough apart to make for a clear distinction between an actual differnce, and measurement error. IOW, the difference could be due to chance.
What Frank is asking for is how statistical analysis was done (there’s liars, damn liars, and statisticians), and what the p-value was (a term that describes how sure you can be whether two groups are actually different from each other, or not). We already know it’s too high (b/c AC stated they were not statistically significantly different), but we don’t know how high it was - just barely, way to high to consider a "tendency to differ’,… FWIW, if this were to be published in a peer-reviewed journal, a statistical section would have to be included.
Frank is also claiming that sample size is 1 (which is, statistically speaking, bad) which is incorrect. It’s just a different level or repetition. E.g., you could test 10 Cervelo frames against 10 frames of another manufacturer, 10 differnt riders on the same two frames, or ride the same bike yourself 10 (or however many) times. It depends on what question you want to answer. In AC’s case, he wanted to know, “on which bike am I faster?”, at least that’s how I read it. No sense testing more than 1 frame each, or different riders.
Makes sense?
FWIW, if every scientific study was done with this much attention to detail, we’d be lucky.
" Where is the statistical analysis?"
Doesn’t reporting the R^2 value and error estimates take care of this?
No
That was helpful. Can someone explain why not.
It is a way of looking at the data and could be useful to determine if the data collection is “valid” but it is not a statistical analysis to let one determine what the chances are the data difference seen is simply due to chance.
So, Frank…
Just curious, but do you go out of your way to rip on everything every well repected expert in the field of bicycle science writes, or is it just anybody who questions the validity of PCs?
Your comments regarding this article of Coggan’s are patently absurd given the nature of the article, and especially so when his summation essentially states that he would not hang his scientific hat on any of it, but that he found it to be personally useful.
Given that he posted this report on Craig’s site, on the forum, which is essentially a group o’ folks chatting about interesting aspects of Bicycle science, the author’s intent certainly doesn’t appear to be to represent this as being in any way authoritative, or anything other than anecdotal.
Too bad Coggan is so freaking methodical that even his BS, just for fun, off the cuff, “interesting note to the guys” stuff reads as being equal science to some of the snake oil being pedalled out there.
I’m not, by the way, accusing you of pedalling snake oil - I don’t have an iron in that particular fire. I am saying that you are carrying around a gigantic axe, and are looking for a place to grind.
Do yourself a favor, and lay off the out-of-context attacks. They are beneath you, and resorting to them cheapens both you and your product.
MH
The problem I have with this article is he (and several ofhis apologists here) have regularly ripped new ones for those who have tried to report similar studies of one done on the PC’s here. He uses his substantial academic credentials to publish such stuff and people believe it even though he knows it has no scientific validity, even though he found it personally useful. Well, the same goes for those who put forth there experiences with PC’s (and other products) here, with I might add, much more substantial differences than the RCH differences reported in this “study”.
Such a “study” or “report” is OK for Coggan to put out on the internet but not for anyone else. It is the height of hypocrisy. It is Coggan and his apologists who carry the axe against the PC’s. I just call it as I see it. You are free to disagree.
Frank
The problem I have with this article is he (and several ofhis apologists here) have regularly ripped new ones for those who have tried to report similar studies of one done on the PC’s here.
What “studies” or “studies of one” on PC’s are you referring to? Post your data, methodologies, and assumptions somewhere.
I’m only aware of the one “study” you always talk about out of the ‘Strength & Conditioning Journal’ which does little to back up your claims.
Everything else you report is of the nature: “I just got an e-mail from a guy…” It makes you sound like the 21st century version of a traveling medicine show.
And… A ‘sample size of one’ studying aerodynamic drag is much more valid than a ‘sample size of one’ studying fitness/physiological changes and adaptations over an extended period of time. You’re talking apples and oranges.
The problem I have with this article is he (and several ofhis apologists here) have regularly ripped new ones for those who have tried to report similar studies of one done on the PC’s here.
What “studies” or “studies of one” on PC’s are you referring to? Post your data, methodologies, and assumptions somewhere.
I’m only aware of the one “study” you always talk about out of the ‘Strength & Conditioning Journal’ which does little to back up your claims.
Everything else you report is of the nature: “I just got an e-mail from a guy…” It makes you sound like the 21st century version of a traveling medicine show.
And… A ‘sample size of one’ studying aerodynamic drag is much more valid than a ‘sample size of one’ studying fitness/physiological changes and adaptations over an extended period of time. You’re talking apples and oranges.
Mr. Curious,
Of course it is not possible to confirm the hypocrisy of AC here and now because he hasn’t posted here, at least under his own name, since the new forum went up. If the old forum was here all one would have to do is search for his responses to many peoples reports here on their experiences with the PC’s. Scientifically, an aerodynamic study of one is no different than a physiologic study of one, despite what you think or say. Even AC did a study of one on RotorCranks and apparently reach some conclusions (so apparently even AC feels it is not out of the range of acceptable to try) yet he has refused to do so for PC’s. Or, do you have a basis for such an absurd statement.
A statistically significant increase in efficiency of 10% in 18 rides over 6 weeks, while not exactly confirming my claims of 40% increase in power in 6-9 months, is not inconsistent with those claims. At least the improvements are statistically significant and published and still AC and his apologists thinks the study to be complete bunk.
I have recently sold a couple of pairs of cranks to people who have told me they intend to keep logs and report their results here. They must have either thick skins or strong stomachs. But, anyhow, if they get bashed about their methodology and lack of scientific rigor (assuming they see improvement) by the AC minions we will just have to remember this thread. I wonder what it will mean if they don’t see “enough” improvement, or any for that matter? We will have to wait and see.
Frank
hello frank,
i’m not flaming you, just genuinely curious, where is the specific study where they find:
A statistically significant increase in efficiency of 10% in 18 rides over 6 weeks.
Thanks,
Shawn T
hello frank,
i’m not flaming you, just genuinely curious, where is the specific study where they find:
A statistically significant increase in efficiency of 10% in 18 rides over 6 weeks.
Thanks,
Shawn T
That is the Lutrell study mentioned here many times and published in the National Strength and Conditioning Journal, Researh in 2003 I think. I have links to the abstract at my web site.
Two groups of trained cyclists one PC group and one reg crank group trained 1 hour 3 times a week in the lab, one on PC’s and one on regular cranks. While the increase in power was not measured, at the same power as before the training the PC group, at the same power dropped their oxygen consumption enough to increase efficiency 10% and their HR dropped 15 beats. The results were statistically significant.