Re: Debunking time again (Re: [CR]1962 Raleigh Gran Sport)

(Example: Production Builders:Teledyne)

From: "Steve Neago" <questor@cinci.rr.com>
To: "Jonathan Cowden" <jac33@tron.arts.cornell.edu>
Cc: <classicrendezvous@bikelist.org>
References: <Pine.3.89.10205311323.A13320-0100000@tron.arts.cornell.edu>
Subject: Re: Debunking time again (Re: [CR]1962 Raleigh Gran Sport)
Date: Fri, 31 May 2002 14:17:00 -0400


Regardless of the outcome, I do not think that a survey of only three cyclists is statistically significant, where we also do not know the test conditions (ie length and duration of the cyclists pedaling on each frame).

Regards, Steve Neago
Cincinnati, OH


----- Original Message -----
From: Jonathan Cowden
Cc: classicrendezvous@bikelist.org
Sent: Friday, May 31, 2002 2:34 PM
Subject: RE: Debunking time again (Re: [CR]1962 Raleigh Gran Sport)



> Hi all. I spoke with a local frame builder about this a while back, and
> he reported the results of a very interesting "blind" study done at
> Cornell with the Cornell cycling team, all of whom were convinced that
> they could tell the difference between different tubesets, up to and
> including Columbus SL and Reynolds 531, which as I understand it, are
> virtually the same. The frames were built with the same measurements and
> painted the same colors. The same components were used in this portion
> of the test. No tubing decals were employed. Frames of
> 531, SL, and mixtures of 531/SL with straight gauge rear triangles were
> compared. Riders took out three bikes: two were the same, one was
> different. They were asked to match the two identical frames. They
> couldn't do it at a rate better than chance, even when one of the frames
> had straight gauge tubing.
>
> Incidentally, this is an elegant demonstration of the importance of blind
> and even "double blind" experimentation. People bring a bunch of
> conceptual baggage to data. In the case of tubesets, the dominant
> conception is that pipes matter independent of anything else. Given that
> mindset, and given the fact that almost everyone knows ahead of time what
> kind of pipes they are riding, should it be surprising that our
> conclusion dovetials with our bias? Only by making the tester blind to
> the pipes can any final conclusions about what testers can and cannot
> discern be determined.
>
> Jon
>
> ps Another interesting finding of the test is that raters are not very
> good at determining the weight of their ride. This is of course implied
> in a not so subtle way about the results I just mentioned. STraight
> guage tubing weighs more, so if riders could detect subtle differences in
> weight, they would have been able to select the frame with the straight
> gauge tubing as "different." These results were extended, however, by
> adding weight to -- if memory serves -- the waterbottles in the form of
> lead. Riders couldn't discern differences of up to three pounds or so,
> even when ascending a hill with a mild incline. Wild.