The Powercrank Thread

[FrankDay]

How does coordination increase power. How would you describe the pedalling of a PC'er on his return to standard cranks.

asked and answered

"In terms of fitness, the use of one leg at a time allows a greater volume of blood flow through the leg per minute (or unit of time) than happens with both legs working together. There is a great mass of muscle in both legs put together and together they can use more blood each minute than the heart can physically supply. When one leg works alone there is plenty of reserve capacity in the heart and a lot of blood can flood into the leg. This increased blood volume may increase the muscle adaptations that are one important outcome that endurance athletes require. The development of larger blood vessel networks in muscle will allow more nutrients to reach all parts of the active muscle more quickly."

first, unless one is having a heart attack the normal heart does not limit our performance. Second, with repeated stress the heart will adapt to have a higher capacity. It is why experienced endurance athletes have higher VO2max than weekend warriors and why PowerCrankers see increases in VO2max with enough training.

 

[CoachFergie]

Endorsed by Chris Froome

http://www.rhinomed.com.au/products/turbine.html

6% more power.

Mythbusters need to do a show on cycling gimmicks.

 

[FrankDay]

Frank, you are doing more than pointing out the study exists and you know it. You are arguing incorrectly that the study has a control group. It does not. You are arguing that the study shows a benefit for training on PCs because of some statistically significant results. It does not. All it shows is a statistically improved performance before and after using PCs. That's all. Because it does not include a control group, there is no way you can tell if this improvement was because of PCs or training effect.

Let's discuss a study that does have a control group, the Burns study, first seen in this thread in post #13 in which they conclude PowerCranks had no benefit.

Burns, JM and Peiffer, JJ and Abbiss, CR and Watson, G and Burnett, A and Laursen, PB, Effects of short-term training with uncoupled cranks in trained cyclists, International journal of sports physiology and performance, 7, (2) pp. 113-120.

Abstract

PURPOSE: Manufacturers of uncoupled cycling cranks claim that their use will increase economy of motion and gross efficiency. Purportedly, this occurs by altering the muscle-recruitment patterns contributing to the resistive forces occurring during the recovery phase of the pedal stroke. Uncoupled cranks use an independent-clutch design by which each leg cycles independently of the other (ie, the cranks are not fixed together). However, research examining the efficacy of training with uncoupled cranks is equivocal. The purpose of this study was to determine the effect of short-term training with uncoupled cranks on the performance-related variables economy of motion, gross efficiency, maximal oxygen uptake (VO2max), and muscle-activation patterns. METHODS: Sixteen trained cyclists were matched-paired into either an uncoupled-crank or a normal-crank training group. Both groups performed 5 wk of training on their assigned cranks. Before and after training, participants completed a graded exercise test using normal cranks. Expired gases were collected to determine economy of motion, gross efficiency, and VO2max, while integrated electromyography (iEMG) was used to examine muscle-activation patterns of the vastus lateralis, biceps femoris, and gastrocnemius. RESULTS: No significant changes between groups were observed for economy of motion, gross efficiency, VO2max, or iEMG in the uncoupled- or normal-crank group. CONCLUSIONS: Five weeks of training with uncoupled cycling cranks had no effect on economy of motion, gross efficiency, muscle recruitment, or VO2max compared with training on normal cranks.

Now one would really think that they compared the PowerCranks group result to the control group result and saw no benefit. But, that is not possible because the control group had a statistically significant drop in efficiency (compared to itself, a la Dixon) such that all the PC group would have to do to show a statistical difference, compared to the control, is have no change in efficiency! Instead the PC group had an increase in efficiency. Further, comparing the control group to the PC group did lead to a statistically significant difference (page 116)

In the NC group, economy (4% decline, P = .01) and efficiency (4% decline, P = .03) significantly decreased from pretraining to posttraining, with a large effect size observed for economy (ES = –1.59) and a moderate effect size observed for efficiency (ES = –1.36). This resulted in a significant interaction in both economy (P = .01) and efficiency (P = .01; see Figure 1) between the groups over pretraining and posttraining time points.

yet Burns simply chose to ignore the control group and determined that the PC group did not show a statistically significant change compared to itself (a la Dixon). Apparently he chose to do this because he couldn't explain the drop he saw in the control group. Rather than use the control group data he had he simply chose to ignore the control group when he did his final analysis because, it appears, he didn't like how it looked. So much for the value of having a control group when it comes to studies.

You guys really need to take a critical reading course. You hold out how awful Dixon is because they don't have a control group (instead, designing the study such that it is a reasonable assumption the control group should hold steady so the participants can act as their own controls) but hold out a study like Burns, with a control group, as being a better study even though it ends up he ignored his control group in his final analysis and ended up doing the exact same analysis Dixon did so he could, apparently, get the result he was looking for.

There is one more thing about the Burns paper. The published paper didn't include all of his data. The paper is 8 papges long. His masters thesis on the same data is over 70 pages long and includes data on how the PC group tested on PowerCranks and on regular cranks. It is clear the intervention was inadequate to actually change pedaling technique because there is a big difference between the two tests, although the gap is closed slightly after the intervention.

If a rider has actually had enough time on the cranks to achieve the intended change there should be no difference in how they test on PC's or regular cranks. That one fact should have been enough for the author to conclude that more training time is necessary to fairly assess the cranks. He, of course, missed this, even removing this data from the published paper.

 

[elapid]

But, that is not possible because the control group had a statistically significant drop in efficiency (compared to itself, a la Dixon)

But wait, didn't Dixon have a control group? Or are you now conceding that you had no idea what constitues a control group and now have finally got a clue?

But, that is not possible because the control group had a statistically significant drop in efficiency (compared to itself, a la Dixon) such that all the PC group would have to do to show a statistical difference, compared to the control, is have no change in efficiency!

No, statistics do not necessarily work that way, Frank. Statistics are not just 1+1=2. Back to the drawing board.

Instead the PC group had an increase in efficiency. Further, comparing the control group to the PC group did lead to a statistically significant difference (page 116) yet Burns simply chose to ignore the control group and determined that the PC group did not show a statistically significant change compared to itself (a la Dixon). Apparently he chose to do this because he couldn't explain the drop he saw in the control group. Rather than use the control group data he had he simply chose to ignore the control group when he did his final analysis because, it appears, he didn't like how it looked. So much for the value of having a control group when it comes to studies.

While Burns et al did not use the normal crank (NC) group as a control group, and that is a pity, he clearly states his methodology in the Methods. He did not have a control group because he was looking at both the uncoupled cranks (UC) and NC groups individually pre- and post-training. He did not ignore the control group, because there was no control group to ignore. He did not decide ad hoc to change his methodology and always intended to compare the UC and NC groups pre- and post-training as per the Methods, and not the UC and NC groups.

However, Williams et al (Cycling efficiency and performance following short-term training using uncoupled cranks. Int J Sports Physiol Perform 4:18-28, 2009) did use two groups of well-trained cyclists with the NC group acting as a control and found no significant differences after 6 weeks of training between the two groups in VO2 max, LT, gross efficiency, or average power during a 30-minute TT. So the only study with a control group shows no difference with PCs.

You guys really need to take a critical reading course. You hold out how awful Dixon is because they don't have a control group

Pot ... met Frank, yet again. We need a critical reading course? Says the guy who has a complete lack of comprehension of study design, scientific methods, and statistical analyses. No, we were saying how awful it was that you did not know what constitues a control group and how awful it was that you use Dixon's study to support PCs when it does no such thing (because it cannot without a control group).

Finally, interesting that you criticize this study for not being long enough (5 weeks) (and no doubt the Williams study), but are fine with studies done by Luttrell & Potteiger (6 weeks) and Dixon (6 weeks) because of their positive findings for PCs. What about Diaz et al's study showing no difference after a 10-week trial period?

 

[FrankDay]

But wait, didn't Dixon have a control group? Or are you now conceding that you had no idea what constitues a control group and now have finally got a clue?

yes, Dixon didn't have a control group. What he had because of his design and experience was the equivalent of or an "effective control group"

By comparing the same patient's numbers before and after treatment, we are effectively using each patient as their own control.

No, statistics do not necessarily work that way, Frank. Statistics are not just 1+1=2. Back to the drawing board.

if you say so. Luckily for me Burns did that calculation and it showed a p-value < .01. Too bad he forgot he did that work when he did his abstract/conclusions.

While Burns et al did not use the normal crank (NC) group as a control group, and that is a pity, he clearly states his methodology in the Methods. He did not have a control group because he was looking at both the uncoupled cranks (UC) and NC groups individually pre- and post-training. He did not ignore the control group, because there was no control group to ignore. He did not decide ad hoc to change his methodology and always intended to compare the UC and NC groups pre- and post-training as per the Methods, and not the UC and NC groups.

??? According to you Dixon's "failure" invalidated his entire study and Burns failure is nothing more than a pity?

However, Williams et al (Cycling efficiency and performance following short-term training using uncoupled cranks. Int J Sports Physiol Perform 4:18-28, 2009) did use two groups of well-trained cyclists with the NC group acting as a control and found no significant differences after 6 weeks of training between the two groups in VO2 max, LT, gross efficiency, or average power during a 30-minute TT. So the only study with a control group shows no difference with PCs.

do you think if someone took a close look at Williams some problems might be uncovered? How does Williams control group differ from Burns NC group that makes his a control group and Burns not?

Pot ... met Frank, yet again. We need a critical reading course? Says the guy who has a complete lack of comprehension of study design, scientific methods, and statistical analyses. No, we were saying how awful it was that you did not know what constitues a control group and how awful it was that you use Dixon's study to support PCs when it does no such thing (because it cannot without a control group).

pot meet kettle indeed. Might want to re-read that excerpt from the Wikipedia article above. LOL

Finally, interesting that you criticize this study for not being long enough (5 weeks) (and no doubt the Williams study), but are fine with studies done by Luttrell & Potteiger (6 weeks) and Dixon (6 weeks) because of their positive findings for PCs. What about Diaz et al's study showing no difference after a 10-week trial period?

Why one 6 week study might get a positive result and another might not should be explainable. We do say that most new users are STARTING to see improvement around 6 weeks such that one might expect a test lasting 6 weeks to sometimes sometimes show benefit and sometimes fail. Further it might depend on the study population also expecting it might be harder to demonstrate a benefit in elite riders than novices. And it might be easier to demonstrate a benefit if an immersion intervention is used vs part-time.

Anyhow, most of these "failed" studies do nothing more than prove what I say is true, if you want to see benefit you need to use them enough and for a long enough period of time. They certainly do not demonstrate the cranks "do nothing" as you hope. It is you who know nothing about study design and interpretation if you insist on trying to hold that these failed studies mean anything regarding what we say about the cranks and pedaling technique.

 

[Tapeworm]

Single legged pedaling does not train the two legged coordination nor does anyone do the drill long enough to train the 5 hour endurance.

Unless they use a counterweight, at which point it depends on how long you can hack being on a trainer. 1.5hr was long enough for me. Regardless, it's still a furphy.

 

[elapid]

yes, Dixon didn't have a control group. What he had because of his design and experience was the equivalent of or an "effective control group

Yes, each patient serves as their own control but there is still not a control group. Because each cyclist served as their own control, all you can say about Dixon's study is that there was an improvement but you cannot ascribe that improvement to any one factor because there is no control group.

According to you Dixon's "failure" invalidated his entire study and Burns failure is nothing more than a pity?

Frank, you are the failure. Again, we are not criticizing Dixon's study per se. He didn't have a control group. That is what it is. What we are criticizing, just to repeat this for the umpteenth time, is your failure to understand a control group and your failure to understand that Dixon's results support PCs because his results cannot do this without a control group.

How does Williams control group differ from Burns NC group that makes his a control group and Burns not?

Critical reading, Frank. Try it. Read the paper, read the methods. It's all in the study design.

yWhy one 6 week study might get a positive result and another might not should be explainable.

It may be explainable for you - 6 weeks is too short for the no difference studies and just fine for the positive studies. LOL. Great scientist you are, Frank. Stick to marketing. At least you have a shot at that for the more gullible consumers.

 

[FrankDay]

Yes, each patient serves as their own control but there is still not a control group. Because each cyclist served as their own control, all you can say about Dixon's study is that there was an improvement but you cannot ascribe that improvement to any one factor because there is no control group.

Now, let me get this straight. Apparently you believe that because each participant has a control that any improvement that any improvement that each participant sees, beyond the control, can be attributed to the PowerCranks? Your only issue now is claiming that the improvement seen by Dixon to PowerCranks is bogus because Dixon analyzed the entire group but because there wasn't a separate CONTROL GROUP his analysis means nothing. Is that correct?

Let's see what Burns described what I mistakenly thought was a control group. From the masters thesis:

The study consisted of a matched-group counterbalanced design, whereby subjects were allocated into one of two groups so that age, body mass and V02max levels were equally matched; that is, the aim was to have no significant differences in these variables between the groups prior to the start of the training interventions. The first group (PC; n = 8) trained with PowerCranks™ while the second group (NC; n = 8) trained with their normal bicycle crank arms .

And from his paper:

This study used a matched-pair design, whereby participants were allocated into 1 of 2 groups matched for age, body mass, training load, and VO2max. An equal number of cyclists and triathletes were represented in each group. Participants were instructed to complete 5 weeks of “regular” training during which 8 participants used traditional bicycle cranks and 8 used uncoupled cranks.

Boy, that sure reads to me like he was trying to get a control group. That fact that he didn't, according to you, have one really shows me I have a lot to learn. Understanding statistics is really tough. I can see now why I am not very good at it. Thanks for pointing that out.

Frank, you are the failure. Again, we are not criticizing Dixon's study per se. He didn't have a control group. That is what it is. What we are criticizing, just to repeat this for the umpteenth time, is your failure to understand a control group and your failure to understand that Dixon's results support PCs because his results cannot do this without a control group.

"Not criticizing Dixon's study?" Could have fooled me! All you are doing is criticizing me for not understanding that Dixon didn't have a control group such that any benefit I attribute to the PowerCranks shows a complete misunderstanding of experimental design and statistical analysis. I guess I must have misread the intent of this sentence by Dixon

In summary, our data suggest that PowerCranks increased maximal aerobic capacity and power in trained cyclists.

Now that you have educated me regarding this stuff and since you are not criticizing Dixon I now know he couldn't have possibly meant that his analysis of his data suggests that PowerCranks actually work. Something else I have misread. I will try to find a reading comprehension class somewhere so I can get better.

Critical reading, Frank. Try it. Read the paper, read the methods. It's all in the study design.

Since, I am so bad at it I was hoping you could do that for me so I could be educated and learn something.

It may be explainable for you - 6 weeks is too short for the no difference studies and just fine for the positive studies. LOL. Great scientist you are, Frank. Stick to marketing. At least you have a shot at that for the more gullible consumers.

Ugh, all one need do is compare the design and methods of the negative and positive studies and see if there is anything different about them that might account for the different outcomes. I have done that (plus, I actually understand the tool being studied) such that I think the differences are perfectly understandable. If you would like to compare any two studies that are seemingly showing different results here I would be happy to do that with you such that maybe you (and the few still following this thread) might learn something. Or, if you are right, maybe I might learn something. Come on, give it a try. You should be able to easily demonstrate to the world how little I know about how to study my own product.

 

[FrankDay]

Unless they use a counterweight, at which point it depends on how long you can hack being on a trainer. 1.5hr was long enough for me. Regardless, it's still a furphy.

How does counterweighted single legged pedaling train the two legged coordination? I am missing something here. Help!

Even if true (which, I doubt, it is), in order to get a 1.5 hour counterweighted work out for both legs would take 3 hours. Not very efficient use of time me thinks.

 

[elapid]

Now, let me get this straight. Apparently you believe that because each participant has a control that any improvement that any improvement that each participant sees, beyond the control, can be attributed to the PowerCranks? Your only issue now is claiming that the improvement seen by Dixon to PowerCranks is bogus because Dixon analyzed the entire group but because there wasn't a separate CONTROL GROUP his analysis means nothing. Is that correct?

No, that is not correct. With pre- and post-comparisons within a single group, all you can say is that there was or was not a significant improvement. You cannot say whether that improvement was due to PCs when you do not compare the PC group to a normal crank group.

Let's see what Burns described what I mistakenly thought was a control group. From the masters thesis:And from his paper:Boy, that sure reads to me like he was trying to get a control group. That fact that he didn't, according to you, have one really shows me I have a lot to learn.

The fact that you have a lot to learn is no surprise to anyone here other than you. Burns had the potential to have a control group, but his study design did not use the normal cranks group as a control group. The only comparison of the two groups was whether the participants in each group were similar or not (so not to have an inexperienced and an experienced group). After confirming that the two groups were similar, all further comparisons were done within the groups (same as Dixon) and not between groups. Critical reading, Frank. Again, you should try it some time.

I will try to find a reading comprehension class somewhere so I can get better.

That would help all of us. Let us know when you've finished your classes. Perhaps we can help you with your homework.

I actually understand the tool being studied ... If you would like to compare any two studies that are seemingly showing different results here I would be happy to do that with you such that maybe you (and the few still following this thread) might learn something. Or, if you are right, maybe I might learn something. Come on, give it a try. You should be able to easily demonstrate to the world how little I know about how to study my own product.

About the only thing you MAY understand is your own product. You certainly do not understand how to read a scientific study or comprehend study design. As previously stated, this is not my area of expertise (not even close). I come here to try and learn. I do, however, know how to critically assess scientific papers and study designs even if they are not within my field of expertise. While I do not have the knowledge or the experience to debate this with you, there are many on this forum who have much more experience than you in this field and none have supported your beliefs. In fact, all have widely lambasted you for outrageous claims.

As far as I can work out, there is only one study which includes a control group and this study showed no difference between normal and uncoupled cranks (Williams et al). This combined with the universal skepticism of PCs by experienced and knowledgeable scientists in this field make all your claims just marketing and nothing else.

 

[FrankDay]

No, that is not correct. With pre- and post-comparisons within a single group, all you can say is that there was or was not a significant improvement. You cannot say whether that improvement was due to PCs when you do not compare the PC group to a normal crank group.

If you say so. Apparently Dixon thought he could. Either way, you cannot say it is not.

The fact that you have a lot to learn is no surprise to anyone here other than you. Burns had the potential to have a control group, but his study design did not use the normal cranks group as a control group. The only comparison of the two groups was whether the participants in each group were similar or not (so not to have an inexperienced and an experienced group). After confirming that the two groups were similar, all further comparisons were done within the groups (same as Dixon) and not between groups. Critical reading, Frank. Again, you should try it some time.

Now help me to understand this. When Burns did a statistical analysis comparing a PC group the the NC group that wasn't the same as comparing to a control group? What need burns do to turn the NC group into a control group, call it a control group?

That would help all of us. Let us know when you've finished your classes. Perhaps we can help you with your homework.



About the only thing you MAY understand is your own product. You certainly do not understand how to read a scientific study or comprehend study design. As previously stated, this is not my area of expertise (not even close). I come here to try and learn. I do, however, know how to critically assess scientific papers and study designs even if they are not within my field of expertise. While I do not have the knowledge or the experience to debate this with you, there are many on this forum who have much more experience than you in this field and none have supported your beliefs. In fact, all have widely lambasted you for outrageous claims.

I see. So, you have critically read Williams and you cannot figure out why it might be negative and Luttrell positive? Williams tried to reproduce Luttrell but looking at much more experienced cyclists, where one might presume change would be harder to come by. Willimas is one of the better papers on this subject as the discussion is pretty good. Perhaps these excerpts from Williams will help you.

The lack of effect of training using uncoupled cranks on GE is in contrast to that of Luttrell and Potteiger.6 A potential reason for these disparate results may be related to the participant recruitment criteria. …

However,
our findings do not discount the possibility that immersion training (completing all training with uncoupled cranks for a period of time) may still elicit improvements …

In conclusion, short-term training (6 weeks) with an uncoupled crank system did not improve Vo2max, LT, GE, or cycling performance in a group of well-trained cyclists. Although more research is needed using longer and more intense training protocols …

As far as I can work out, there is only one study which includes a control group and this study showed no difference between normal and uncoupled cranks (Williams et al). This combined with the universal skepticism of PCs by experienced and knowledgeable scientists in this field make all your claims just marketing and nothing else.

I suggest you read Williams more carefully again. All Williams proves is that 6 weeks of part-time training is unlikely to improve performance variables in well-trained cyclists to a statistically significant detail. I actually agree with that conclusion. Even Williams recognizes his study is very narrow and doesn't really prove much. What is wrong with you?

 

[elapid]

Apparently Dixon thought he could. Either way, you cannot say it is not.

Exactly. But you cannot say to everyone here that PCs make a difference on the basis of Dixon's study because the design does not allow you to say it was or was not PCs.

When Burns did a statistical analysis comparing a PC group the the NC group that wasn't the same as comparing to a control group? What need burns do to turn the NC group into a control group, call it a control group?

Burns needed to compare the results of the two groups. He did not do that. He just compared the pre- and post-training results within the two groups, and not between the two groups.

I see. So, you have critically read Williams and you cannot figure out why it might be negative and Luttrell positive?

No, you don't see. I didn't say any such thing. All I said is that YOU reject studies which find no significant results because they are too short and you accept studies which show significant results despite them being the same length of time.

I suggest you read Williams more carefully again. All Williams proves is that 6 weeks of part-time training is unlikely to improve performance variables in well-trained cyclists to a statistically significant detail. I actually agree with that conclusion. Even Williams recognizes his study is very narrow and doesn't really prove much. What is wrong with you?

What is wrong with you? All I said is that the Williams et al paper is probably the best paper on PCs because it has a control group. None of the other studies, regardless of their findings, has a control group. Comprehension skills? When are you taking your class?

 

[FrankDay]

Exactly. But you cannot say to everyone here that PCs make a difference on the basis of Dixon's study because the design does not allow you to say it was or was not PCs.

I don't have to say PC's made a difference in the Dixon study because Dixon said it. While you may not believe he should do so apparently he (and his co-authors) believed their data deserved such a conclusion. Otherwise why would they say it?

What need burns do to turn the NC group into a control group, call it a control group?

Burns needed to compare the results of the two groups. He did not do that. He just compared the pre- and post-training results within the two groups, and not between the two groups.

Wrong. First you tell us he didn't have a control group, all we had to do was read his methods to understand this. Now you tell us that to have a control group all he need do was to compare the NC group to the PC groups, which you believe he didn't do. Speaking of poor critical reading skills, you must have missed this

In the NC group, economy (4% decline, P = .01) and efficiency (4% decline, P = .03) significantly decreased from pretraining to posttraining, with a large effect size observed for economy (ES = –1.59) and a moderate effect size observed for efficiency (ES = –1.36). This resulted in a significant interaction in both economy (P = .01) and efficiency (P = .01; see Figure 1) between the groups over pretraining and posttraining time points.

He did it. He did what you said he needed to do. He just failed to mention this statistically significant finding in his abstract or conclusion.

No, you don't see. I didn't say any such thing. All I said is that YOU reject studies which find no significant results because they are too short and you accept studies which show significant results despite them being the same length of time.

I don't reject any study. All I have done is point out the weaknesses of some of the studies that people are interpreting as "proving" my product worthless. To prove the product worthless it would seem to me, at a minimum, a study needs to involve enough intensity and enough time (appropriate to the studied population) that anecdotal experience suggests is necessary to see a difference. While a study may occasionally luck out (come close to the bare minimum to see a difference) and get a positive result (Luttrell, Burns, Dixon) those studies hardly examine the full potential of the device.

What is wrong with you? All I said is that the Williams et al paper is probably the best paper on PCs because it has a control group. None of the other studies, regardless of their findings, has a control group. Comprehension skills? When are you taking your class?

LOL. Here is what you wrote regarding Williams

So the only study with a control group shows no difference with PCs.

as if that finding was somehow definitive. Even Williams doesn't say that. Williams should have been negative because the study design was so weak no one who knows anything about the PowerCranks would have ever predicted that such an intervention would achieve a positive result in that cohort. If Williams had bothered to talk to me about his design before starting I could have saved him the effort. As I said, at least he "proved" that if you want to see a benefit in that time frame you had better follow the Dixon example rather than the Williams example.

Further, it appears that now you would agree that Burns had a control group because he did compare groups and the result rejected the null hypothesis. Edit: In addition, Luttrell had a control group and found a positive difference. So, the statement that the only study with a control group found no difference is not correct.

 

[elapid]

Frank, you are good for a laugh but very little else. I am taking a break from your insanity. If you want to twist everything I say then go ahead and troll away. I've had enough of it, but no doubt will come back in a couple of months to see that the soap opera has not changed tunes. If you want to cherry pick studies which show a difference and ignore studies that don't, then everyone here knows your BS. None of the studies support the sort of claims you are making, even if they show a positive result. None of the scientists and researchers in this field on this forum believe your claims can be true, so why should I believe a septuagenarian with no training in this field, no scientific writing/reading/designing ability, and a history of very questionable ethics with marketing over substance? See you in a couple of months, Frank. I hope you have sat your comprehension class by then and perhaps even a scientific writing course as well. May be we can have more intellectual debates with you by then rather than just a written form of who can shout the loudest and longest. The latter you win hands down, I have better things to do with my time.

 

[FrankDay]

If you want to twist everything I say …

Pot, meet kettle, YOU!!! See you around. Come back when you are willing to discuss and eschew name calling.

 

[FrankDay]

None of the studies support the sort of claims you are making, even if they show a positive result.

Of course they don't. Our "claims" require exclusive use for 6-9 months. When a study is done that includes those requirements in an "average" racer then we will see if the claims are supported or not. LOL

 

[CoachFergie]

Frank, you are good for a laugh but very little else. I am taking a break from your insanity.

Just ignore him and report people when they spam this and other forums.

 

[Alex Simmons/RST]

Of course they don't. Our "claims" require exclusive use for 6-9 months. When a study is done that includes those requirements in an "average" racer then we will see if the claims are supported or not. LOL

Yet your own website's FAQ says:

How soon will improvements become noticeable?

In most sports most users have reported seeing running and balance improvements starting within 2 weeks. For cyclists measurable improvements will take approximately 4-10 weeks.

emphasis is mine.

So either a difference is detectable in relatively short time frame or it's not.

Which is it Frank?

 

[FrankDay]

Yet your own website's FAQ says:

How soon will improvements become noticeable?

In most sports most users have reported seeing running and balance improvements starting within 2 weeks. For cyclists measurable improvements will take approximately 4-10 weeks.

emphasis is mine.

So either a difference is detectable in relatively short time frame or it's not.

Which is it Frank?

Ugh, just because a few people see improvement starting in 4 weeks of immersion training is not a good reason to design a study lasting 5 weeks of part-time training and expect to see a significant difference. Statisitical significance in the scientific sense is a lot different than "detectable". LOL.

If one wants to design a study that should definitely show a significant difference and one wanted to listen to us it seems one should design a study lasting 10 weeks involving immersion training. That way one would, it seem, pick up almost all of the early improvement according to our "claim." Dixon managed to show that difference in 6 weeks but, if his cohort had been different would it have meant much if he had failed? Probably not.

The fact that Elapid and Coggan and you are internet bullies does not mean that you get away with nothing more than opinion where no one else can. I will call you on it which is why I am held in such disdain in your circle. My proof of my belief may be lacking but your proof that I am wrong is, simply, non-existent.

 

[Alex Simmons/RST]

Ugh, just because a few people see improvement starting in 4 weeks of immersion training is not a good reason to design a study lasting 5 weeks of part-time training and expect to see a significant difference. Statisitical significance in the scientific sense is a lot different than "detectable". LOL.

If one wants to design a study that should definitely show a significant difference and one wanted to listen to us it seems one should design a study lasting 10 weeks involving immersion training. That way one would, it seem, pick up almost all of the early improvement according to our "claim." Dixon managed to show that difference in 6 weeks but, if his cohort had been different would it have meant much if he had failed? Probably not.

The fact that Elapid and Coggan and you are internet bullies does not mean that you get away with nothing more than opinion where no one else can. I will call you on it which is why I am held in such disdain in your circle. My proof of my belief may be lacking but your proof that I am wrong is, simply, non-existent.

I'm not a bully. I just call BS when I see it. Which happens pretty often with your posts.

If you don't write nonsense you won't get called out for it. Pretty simple really.

 

[Tapeworm]

P values:-
http://m.nautil.us/issue/4/the-unlikely/sciences-significant-stats-problem

 

[FrankDay]

I'm not a bully. I just call BS when I see it. Which happens pretty often with your posts.

Sure you are a bully. Anything you disagree with you consider BS even though you may have nothing more than your gut feeling in support of your view. You can't have a reasonable discussion regarding disparate views without resorting to personal attacks if the other side doesn't come around to your view. That, in my book, qualifies you as a bully. That may "work" as a debating tactic (i.e., they go away) with many but, as you may have discovered, doesn't work well with me.

 

[FrankDay]

P values:-
http://m.nautil.us/issue/4/the-unlikely/sciences-significant-stats-problem

Yep. Statistics is tough. Few here (including me) are really good at it. Most of us have to trust that the statistics reported in most studies was chosen properly and done correctly.

No single study can ever prove anything because statistics, even if properly applied, can only give the probability that the data represents random difference or real difference. p<.05 is simply a convention generally accepted as representing a large enough probability (>95% chance of a real difference) to be referred to as "significant". There is still a chance that the conclusion is not correct just as a p<0.1 doesn't qualify, per the convention, as being a significant difference but there is still a 90% chance the data represents a real difference. And, all this assumes the researcher has chosen the correct analysis for the data. Edit: This is why studies need to be repeated and confirmed (sometimes many times) before one can feel really comfortable that the findings are really true.

P-values are nothing more than a statistical tool to help people put research into perspective. Putting too much into the results of a single study can lead people down the wrong path. That could be the case here with PowerCranks if you put too much trust in those poorly designed "negative" study's or too much trust in a single positive study.

Regardless of what the studies show, I would still think our average customer would most value our money-back guarantee because what they are most interested in is whether the tool works for them, not what the tool did in some study.

 

[CoachFergie]

P values:-
http://m.nautil.us/issue/4/the-unlikely/sciences-significant-stats-problem

And...

http://evidencebasedfitness.net/why-not-0-06/

http://evidencebasedfitness.net/this-episode-is-brought-to-you-by-the-letter-p/

 

[FrankDay]

And...

http://evidencebasedfitness.net/why-not-0-06/

http://evidencebasedfitness.net/this-episode-is-brought-to-you-by-the-letter-p/

My goodness, Fergie has actuallly posted something useful and on topic, at least as far as the topic has veered this way, without any snide remarks. Progress.

My guess is the reason 0.05 has become the de facto p-value of choice by publications is it is very hard to get a smaller p-value such that journals would have a lot of trouble finding enough studies with positive results to publish if 0.01 were the standard. If they chose 0.1 there would be way too many papers with positive results. Hence, it is a useful choice for those who determine what gets published and what doesn't.