As many of you rec.bicycles readers are aware, there have been occasional,
sometimes acrimonious, discussions about how some frames are so much 
stiffer than others.  Cannondale frames seem to take most of the abuse.
The litany of complaints about some bike frames is long and includes
excessive wheel hop, numb hands, unpleasant ride, broken spokes, 
pitted headsets, etc.  I was complaining to a friend of mine about how there
was so much ranting and raving but so little empirical data - to which
he replied, "Why don't you stop complaining and do the measurements
yourself?".  To that, I emitted the fateful words, "Why not, after all,
how hard can it be?".  Following some consultation with Jobst and a few
other friends, I ran the following tests:

The following data were collected by measuring the vertical deflection at
the seat (ST), bottom bracket (BB) and head tube (HT) as a result of 
applying 80lb of vertical force.  The relative contributions of the 
tires, wheels, fork, and frame (the diamond portion) were measured using 
a set of jigs and a dial indicator which was read to the nearest .001 
inch.  For some of the measures, I applied pressures from 20 to 270 lbs 
to check for any significant nonlinearity.  None was observed.  The same 
set of tires (Continentals) and wheels were used for all measurements.  
Note that these were measures of in-plane stiffness, which should be 
related to ride comfort, and not tortional stiffness which is something 
else entirely.

Bikes:

TA - 1987 Trek Aluminum 1200, this model has a Vitus front fork, most 
     reviews describe this as being an exceptionally smooth riding bike

SS - 1988 Specialized Sirus, steel CrMo frame, described by one review as
     being stiff, hard riding and responsive

DR - 1987 DeRosa, SP/SL tubing, classic Italian road bike

RM - 1988 Cannondale aluminum frame with a CrMo fork, some reviewers
      could not tolerate the rough ride of this bike


               TA              SS              DR              RM
           ----------      ----------      ----------      ----------
	   ST  BB  HT      ST  BB  HT      ST  BB  HS      ST  BB  HT
diamond	    1   1   0       2   2   0       2   2   0       1   1   0
fork        3  11  45       3   9  36       4  13  55       3  10  40
wheels      2   2   2       2   2   2       2   2   2       2   2   2
tires      68  52  66      68  52  66      68  52  66      68  52  66
total      74  66 113      75  65 104      76  69 123      74  65 108


What is going on here?  I read the bike mags and this net enough to know 
that people have strong impressions about the things that affect ride 
comfort.  For example, it is common to hear people talk about rim types 
(aero vs. non-aero), spoke size, butting and spoke patterns and how they
affect ride.  Yet the data presented here indicate, just a Jobst predicted,
that any variation in these factors will essentially be undetectable to 
the rider.  Similarly, one hears the same kind of talk about frames, 
namely, that frame material X gives a better ride than frame material Y, that
butted tubing gives a better ride that non-butted, etc.  (I may have even 
made such statements myself at some time.)  Yet, again, the data suggest 
that these differences are small and, perhaps, even undetectable.  I offer
two explanations for this variation between the data and subjective reports
of ride quality.

Engineering:
These data are all static measurements and perhaps only applicable at the
end of the frequency spectrum.  Factors such as frequency response, and 
damping might be significant factors in rider comfort.

Psychology:
There is no doubt that these bikes all look very different, especially the
Cannondale.  They even sound different while riding over rough
roads.  These factors, along with the impressions of friends and reviews
in bike magazines may lead us to perceive differences where they, in fact,
do not exist.

Being a psychologist, I am naturally inclined toward the psychological
explanation.  I just can't see how the diamond part of the frame contributes 
in any significant way to the comfort of a bike.  The damping of the frame 
should be irrelevant since it doesn't flex enough that there is any
motion to actually dampen.  That the frame would become flexible at
some important range of the frequency spectrum doesn't seem likely either.

On the other hand, there is plenty of evidence that people are often very
poor judges of their physical environment.  They often see relationships 
where they don't exist and mis-attribute other relationships.  For example,
peoples' judgement of ride quality in automobiles is more related to the
sounds inside the automobile than the ride itself.  The only way to get
a good correlation between accelerometers attached to the car seat and
the rider's estimates of ride quality is to blindfold and deafen the
rider (not permanently!).  This is only one of many examples of mis-
attribution.  The role of expectation is even more powerful.  (Some even
claim that whole areas of medicine are built around it - but that is 
another story entirely.)   People hear that Cannondales are stiff and,
let's face it, they certainly *look* stiff.  Add to that the fact that
Cannondales sound different while going over rough roads and perhaps
the rider has an auditory confirmation of what is already believed to
be true.  

Unless anyone can come up with a better explanation, I will remain 
convinced that differences in ride quality among frames are more a 
matter of perception than of actual physical differences.