[CR]frame flex: experiments and experience

Ted Ernst wrote:

If we agreed on everything it would be dullsville. My experience works for me and your knowledge works for you. That's good and turns the hubs of hell to heat us up once in a while.

Ted,

I agree that differences in perspective are what make CR so interesting. However, your "you have your view and I have mine," attitude is almost inadvertently antagonistic in this particular case. The analytical engineering folk are presenting arguments that they "know" to be more than just their view. They are putting forth what they see as the truth. It's like when someone tells me that it's okay for me to believe in organic evolution, but they'll take their creationism, thank you very much. NO, THAT'S MISSING THE POINT; I'M RIGHT AND THEY'RE WRONG. I'm talking scientifically verifiable theory, they are talking fantasy. In other words, in my view, my view is not just my view.

Of course, in this "frame feel" case I actually don't believe the staggeringly weak analytical data put forth are anywhere near enough to support some of their arguments. My particular favorite is this little gem, excerpted from Bob Bundy’s summary of his frame flex experiment:

“Yet the data presented here indicate, just a Jobst predicted, that any variation in these factors will essentially be undetectable to the rider.”

Wha, wha, wha???? Am I missing something here? Talk about being influenced by expectations! How did anyone make the leap from the data to an affirmation of Jobst’s predictaions? Because the data don’t “look like” they vary by “all that much” from bike to bike? There is absolutely ZERO statistical analysis, or any other analysis, of the deflection measurements. There are no replications. This experiment isn’t worthy of a “B” in a high school lab science class. How is it that an 18% increase in deflection at the headtube between the Specialized and the DeRosa can be dismissed as “essentially undetectable” ? Because the units are in thousandths of an inch, and that’s “real small”? There is a huge unanswered question here. If we assume that this grossly oversimplified experiment somehow represents that dynamic realities of a bike on the road, and that the measurements are precise, accurate, and repeatable, how do we go from “thousandths of an inch” to what a rider will feel? I find it a little scary that a psychologist could exhibit such a weak grasp of experimental technique and particularly of the importance statistical data evaluation. Don’t even get me started about Jobst.

Warning, O/T analogy coming…This whole mess reminds me of the arguments between high-end stereo guys. The guys from audio magazine A say that any stereo receiver with low harmonic distortion and flat response, as measured with instruments, will sound good and you don’t need to listen to the gear to know that you’ll be getting good sound. They give so little credit to the human being and his/her abilities of perception that it saddens me. On the other side are the guys from audio mag B, who say that none of the measurements tell you anything and that it’s all in the listening, yet they don’t do even single-blind comparisons. Personally, I find magazine B more interesting because they take the more complex view, look at so many more factors, and in the end they bring it all back to a human experience. But boy do I wish they would put some weight behind their subjective evaluations with some blind tests. But, paraphrasing Mr. Maas, people don’t want to feel like their judgment can be called into question. As experts they feel that they are totally objective when making subjective evaluations. Funny.

Tom Dalton

Bethlehem, PA

---------------------------------
Start your day with Yahoo! - make it your home page