RISK COMPENSATION

[slowster]

With regard to the argument that risk compensation would negate the benefit of using a mirror, the following quote by CJ is arguably an example of just that:

NEVER ride into low sun on a high-speed road without a mirror.

(viewtopic.php?p=780968#p780968)

[Cugel]

And, if one doesn't sometimes attempt to study things rigorously, to extract some repeatable facts, it’s all too easy to be guided by various sorts of misunderstanding and misinterpretations based on the limits of personal observation and personal deductive reasoning.

Formal study has its limits. Informal study has its limits.

The problem with "extracting the facts" is that facts can be rather too easily formed, as they're extracted, into what the extractor wishes to shape. Far too many "studies" are poorly constructed and composed with a hidden agenda demanding that they reach a predetermined result, producing thereby "facts" to fit.

Of course, less structured observations of happenings formed into anecdotes from various personal observations can suffer the same problem. We all have our epistemological beliefs and assumptions In either case (rooting out cause and effect formally or informally) there needs to be some attempt to avoid the various prejudices or desires of those attempting to find "the facts". In so many cases, this can be very difficult to do, especially when a matter is very complex, involving a multiplicity of factors that may be causes or may be co-incidences. Even worse when the matter has become ideological bordering on religious.

Yet we humans do have some form of intelligence that allows us to learn the hard lessons about various risks. Many performing risky antics learn that it's all too easy to over-estimate the degree of "safety" provided by a device supposedly designed to increase some aspect of safety. Motoring is full of them. I must have heard a hundred times, when they first became available, how a driver was convinced they could drive much faster because, "I have ABS brakes now". Of course, ABS brakes do nothing to reduce the driver's reaction time to alarming events appearing rapidly through the windscreen.

If you look through your own life, I feel sure you'll have noticed endless examples of similar ilk, concerning all sorts of safety devices, procedures and so forth. "I've got one of these now so I'm completely safe from all possible harms".

[Nearholmer]

The first paragraph appears to be a long way of asking to return to pre-scientific modes of thought and decision-making.

No thanks.

As regards my own life, I can think of instances of safety interventions which have been undermined by people taking more risks, and safety interventions where that hasn’t happened, so they’ve worked well, and safety interventions which have fallen flat on their face because people have decided that they are more trouble than they are worth, so have ignored or sabotaged them.

I can also think of instances where safety measures were resisted for years, decades in fact, on the basis that they would encourage people to adopt more risky behaviour, and when they were eventually adopted, people didn’t do that at all.

[pjclinch]

Whether RC applies to any particular safety measure in any particular activity, and if so to what extent, is guesswork unless or until demonstrated. As I keep saying, there is no demonstration that it applies as a general rule, or that where it does apply it always has a significant effect.

It's a general human behavioural characteristic to change behaviour with respect to perceived risk. The more you look at everyday interactions the more obvious this is. That there are people who won't cycle without a helmet is a clear example of how helmet interventions impact risk taking in cycling.

There is no demonstration it is always a significant effect on safety, but the same could be said of speed: that doesn't mean we stop worrying about speed when trying to make things safer (though as a lot of high speed prangs show a lot of people try). Sometimes negligible safety effect will be down to the physical situation (I can take very few risks going up a big hill at a very relaxed pace that doesn't force me to zig zag, I can take lots coming down the other way), sometimes it will be down to the cultural outlook and personality of the rider, but it is a general rule of human behaviour that behaviour changes with perceived risk.
I've never met you in person, I've never seen you ride a bike, but I'd wager a couple of pints at least that if I met you on a stretch of road you're very used to and we swapped bikes for a mile, you taking my recumbent and me taking your upright, that you would be much more worried and careful on the unfamiliar bike. You're not used to it yet, you can't relax, you become more vigilant, you take more care. You feel more at risk of crashing and your behaviour changes. This is easy to predict as a general principle because I've seen quite a few folk (but not a representative statistical sample of All Cyclists, by any means) go through that, been through it myself and noticed it doubles for any situation where I'm using something outside my ability to relax. Where something puts me inside my ability to relax (possibly equipment, possibly skill/experience, possibly a combination) then I lower my vigilance and perceived risk aversion (not necessarily absolute risk though, there's plenty of stuff that is actually safer, particularly off-road, if you attack it with a sense of "I'll cruise through this" rather than tentatively have a go having decided it's likely to go wrong).

Where the uncertainty creeps in is exactly when and how it will cut in for any particular person in whatever context. I take more risks of I'm wearing a helmet on my MTB doing formal trails than if I don't: I quite deliberately push myself more with the special hat, and make sure I wear gloves too. But I would be very surprised if there'd be any tangible difference in the risks I take helmet or not on my Saturday morning trundle to the bakers. I'm not in a hurry, it's all about getting the croissants, there really isn't much scope for making things riskier in the context of that ride, while there's lots on an MTB trail. My bakery run with probably no RC does not disprove RC in cycle helmet use, it's just other factors in the journey are far more significant than the (very low) risks involved.

And there are further spanners in the works because we're dealing with perceived risk, not actual risk. For example a lot of folk ride in the gutter because they're too scared to move out, by providing the confidence of more perceived safety provided by a lid they're better able to move out and that (rather than the direct benefit of head protection) can make them safer. And so on.

But that people tend to act differently when they feel tangibly safer to the point that they can relax more really isn't in doubt. That you're asking for a specific formal proof for each specific context doesn't really make much sense.
"Okay, the waiter takes more care walking from the bar to my table with a brim-full glass of lager than a whisky. I wonder if he'll take more care with a brim- full glass of stout too, or take more care with two full bowls of soup than the sandwich he took over earlier, because I haven't seen him carry those yet so how he'd behave could only be a guess"

RC annoyingly gives you two bad options: pretending it doesn't happen because it's inherently hard to predict, or adding error bars where you don't know how big they are to aspects of safety where human behaviour can have a significant effect on the outcome. Trains are very safe because most of the scope for human behaviour has been removed. That's much harder to do with bikes.

That some people over-claim with considerable certainty that RC wipes out any benefit of helmets is true, but there are always folk ready to grasp the wrong end of any given stick, much as those absolutely certain that helmets prevent 85% of serious head injury because it said so in a reputable peer-reviewed journal.

Pete.

[pjclinch]

Far too many "studies" are poorly constructed and composed with a hidden agenda demanding that they reach a predetermined result, producing thereby "facts" to fit.

"Policy based evidence making" is the term in government/civil service circles, actually used in formal documents advising not to do it IIRC. Not that that advise is always heeded, mind. It happens in academic papers too, because at the end of the day scientists are only human.

Pete.

[Cugel]

The first paragraph appears to be a long way of asking to return to pre-scientific modes of thought and decision-making.

No thanks.

In many domains of consideration, study, fact-finding and procedure-inventing the scientific method is the best method humans have so far invented. But, as you know, this method requires feedstock of the more resilient kind of facts, as in physics and chemistry rather than, say, those akin to the assumptions of astrology or phrenology.

Many domains of consideration do not posses a resilient and well-proven feedstock of facts but rather a vast collection of long-extant human prejudices, assumptions, beliefs, desires and other base notions that have little to recommend them as "facts".

These domains of consideration include subject matters such as psychology, sociology, anthropology and economics. Modern practitioners like to claim their studies in these domains as "sciences" but they are in effect more SSOs (Science Shaped Objects)".Scientific procedures may be followed but the feedstock of "facts" employed in the study of their relationships of cause & effect is often an entirely artificial feedstock of very plastic beliefs and assumptions wrought into "facts" that already contain the answers about cause and effect preferred by the scientists.

Studies concerning cycling helmets are usually of these inchoate and highly artificial constructs. The most cogent thing they reveal is the pre-existing attitudes of those constructing the study.