On Aug 6, 6:53 pm, Tim McNamara <tim...@bitstream.net> wrote: > While Jan et al made a lot of efforts to minimize the variables- the goal of > experimental design being to eliminate all fariables except the one you want > to measure- doing a coast down ride with a live human on board, outside, > etc., has too many variables that may intrude. They worked hard but the > design of the experiment doomed it to having too much noise, even with > careful statistical analysis, to produce reliable and valid results.
The statistical analysis is a tool to determine how much "noise" you have. If the same tires always score the same in repeat experiments, but other tires always score differently, then you have shown that the tires are different. If there is too much noise, you will find that the differences between repeat tests of the same tire are greater than the differences between different tires. The statistics are just a formalized way of evaluating that. So to say that "the results are flawed despite the statistical analysis" doesn't make sense. What it comes down to is both the amount of external variables and the size of the differences in tire performance. In an extreme example, if you measured a square wheel vs. a round one, you could have all kinds of wind and other variables, yet you would show conclusively that the round wheel is faster. (The square one doesn't turn at all.) What surprised us with our testing was that the performance differences between tires are huge - large enough to measure even in a test that has some (albeit small) amount of noise. I had hoped that with the recent trend of even pro racers to wider tires at lower pressures, the tire discussion was finally over. Yes, we did use methods that were different from the industry standard. However, the industry (at least in North America, where Bicycle Quarterly is mostly read) has since accepted our results, which means that a) they found our methods credible and b) they were able to replicate our results with whatever testing methods they chose. Jan Heine Editor Bicycle Quarterly http://www.bikequarterly.com -- You received this message because you are subscribed to the Google Groups "RBW Owners Bunch" group. To post to this group, send email to rbw-owners-bunch@googlegroups.com. To unsubscribe from this group, send email to rbw-owners-bunch+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/rbw-owners-bunch?hl=en.
