You've misapplied the burden of truth. It is NOT on helmet skeptics. The burden of proof is upon those claiming safety benefits for helmets.

Okay, this is a pointless three-ring circus, but I have opinions! So I'll share 'em. Keep in mind: I regularly wear a helmet (on road rides) and am NOT anti-helmet.

1. The only studies that have been done on helmets (because they are the only practical ones) are population-based statistical analyses, which attempt to measure incidence of head injury in the population before and after the passing of helmet legislation, and case-control studies which compare the average degree of head injury severity between cyclists who had been wearing a helmet and those who had not. These are not super-reliable studies, because there are any number of confounding variables that cannot be accounted for. So that's one problem. In any case, the statistical studies generally show no benefit to helmets, while case-control studies generally show some benefit. No matter which kind of study you look at, the results are inconclusive. My interpretation of the general results from each kind of study is that helmets probably make little difference on a population basis, but may save your bacon in the event that you are actually in an accident with a serious head impact. Maybe.

On a related note, there is some possibility that helmets could actually increase the risk of certain kinds of head trauma. It is torsional forces that are the real killer when it comes to head injuries, and helmets may make it more likely for an impact to result in severe torsional forces on the brain. The tendency of helmet manufacturers to put all kinds of doo-dads on their helmets and to mold them in non-round shapes really bothers me. A helmet should be a roundish dome! Pointy streamlined bits coming off the back of a helmet constitute a safety risk. I'll happily make an exception for TT helmets, though.

2. The plural of anecdote is not data. No one should be impressed by "my helmet saved my life!" stories. Well, maybe it did. And maybe it didn't. There are way too many variables in any accident, including ones that are severe, to draw any kind of consistent conclusion about the efficacy of helmets. I don't care if you couldn't see straight or were knocked unconscious or got a concussion or what. It's not research, and it's not data.

Think about this one. It seems that 1 in every 10 cyclists has a story about how a helmet has saved his or her life/saved them from serious head injury. There's simply no way to take this seriously - the sheer volume of these anecdotes would have us convinced that, before helmets arrived, cyclists were dropping like flies. Either there is an unambiguous, statistically significant difference in the number of head injuries received by cyclists in, say, 1985 versus 2005, or most of the people walking around who think that their helmet saved them from serious injury or death are mistaken. The former is not the case.

3. Related to #2. I am amazed that I continue to see the broken helmet myth trotted out here on BF, though I suppose I shouldn't be. It's been explained a million times that a cracked helmet has failed. Helmets work by crushing. The foam does NOT absorb energy through brittle failure. Cracking is an undesirable failure mode that reduces the effectiveness of your helmet. A cracked helmet is a helmet that has not done its job well.

4. Cognitive dissonance is not grounds for dismissing the passing study linked by the OP. True, the results doesn't jibe with what most of us "know" about helmets. That does not make it invalid! It's not a perfect study, but that 3.5" difference has been tested, and it certainly is statistically significant. I also happen to agree that the fact that the scientist was hit twice while wearing a helmet has little bearing on the results - being hit twice is NOT statistically significant. What does this imply about your risk when wearing a helmet vs. not wearing a helmet? Uh, not a lot. Drivers give less passing distance when you wear a helmet, which probably slightly increases your risk of being struck by a vehicle. Okay, so what? That doesn't tell us anything about whether a helmet is effective in preventing injury or not. Where's the controversy?

Anyway, like I said, I wear my helmet. I figure it's cheap insurance that probably won't hurt. But choosing not to wear a helmet is far from stupid, on the evidence. We all like to think that we're smart and reasonable people, but helmets have little to do with rationality.

Now, from this point further I will refrain from any more beating of this dead horse .