Do measurements of motorcycle conspicuity really measure conspicuity effects?

Accident studies have indicated that lack of detectability is a factor in daytime multivehicle accidents involving motorcycles. Until recently, little research had been conducted using direct behavioural observation procedures, but this has now changed with the procedures used involving gap-acceptance, detection and recall. Several problems with the use of these procedures to measure effectiveness of conspicuity ‘enhancers’ are outlined. It is concluded that the gap-acceptance procedure should not be used for this purpose; detection studies need to control for context effects, especially that of expectancy, and also to hold luminance contrast constant from the real scene to the simulation; recall studies need to investigate the influence of memory-related factors on recall performance, in addition to using moving motorcycles, rather than stationary motorcycles as in the studies to date.