Horse wrote:There's a clear progression, for example the typical 'SMIDSY' situation:
Driver arrives at junction . . .
1 Doesn't look
2 Looks, but rider is obscured (eg car door pillar, street furniture, other vehicle) [ie 'not visible']
3 Looks, but doesn't notice the rider
4 Looks, notices the rider, but makes incorrect decision on rider's speed and distance
5 Looks, makes correct decision about rider's speed and distance, but pulls out anyway because driver doesn't understand limitations on rider's braking ability etc
6 Looks, correct decisions, pulls out because they don't care
Hi-viz can really only influence '3' - and we're dealing with situations where the rider is close to the junction anyway!
Agreed but even for '3', there are:
3.a. Looks, expects clear road and 'sees' that image. Hi-viz might (to be proved) cause an adjustment of that assumption; But drivers also pull out on large vehicles like this - even those wearing hi-viz. Images generated by planning (in the same part of the brain as process vision) are not always corrected by input, particularly where that input is easily assumed to be something else expected (speculatively: such as road signage).
3.b. Looks, sees much traffic large vehicle traffic, smaller rider looks like a gap. Again, hi-viz 'might' cause readjustment, or might be unconsciously dismissed ('cos brains work that way
) as 'not a threat', or might be camouflaged (as Michael Mason was speculated to be, in a different scenario) against a complex background with many light sources.
All in all, a consideration of where hi-viz might help and how, does appear to show why there never seems to be evidence of a reduction in injury rates in population-level statistical analyses. It's potentially helpful actions are too easily over-ridden by other considerations.