Horse wrote:waremark wrote: I don't know what category of accident this graph is supposed to measure
You'd probably need to delve into the original research, e.g. "1 The effects of age and experience on collision risk, from Maycock et al. (1991; figure reproduced from
Maycock, 2002)"
https://trl.co.uk/sites/default/files/TRL527.pdfNovice driver accidents and the driving test
Prepared for Road Safety Division, Department for Transport, Local Government and the Regions
G Maycock
TRL Report TRL527
First Published 2002
ISSN 0968-4107
3.3.2 Self-report accident liability studies
Accident liability survey
Figure 4 shows the modelled results of an accident liability survey (Maycock, Lockwood and Lester, 1991) based on accidents reported by a sample of about 13,500 drivers split roughly equally between the sexes; the accidents were self-reported accidents on public roads, and were mainly damage only. The study was a cross-sectional study using a structured sample in which older and younger drivers, and drivers with limited driving experience were deliberately over-sampled. In considering the relationships shown in Figure 4 therefore, it should be borne in mind that the modelled effects are primarily ‘between driver’ effects which may or may not be precisely the same as a longitudinal ‘within driver’ effect – i.e. how accident rate changes for an individual driver as he or she ages and gains experience.
With this caveat in mind, the upper broken lines in Figure 4 show how the accident liability of male and female drivers in their first year (i.e. novice drivers) falls with increasing age.
…
In addition to the issue already mentioned of ‘cross-sectional’ versus ‘longitudinal’ effects, there are a number of other factors that need to be borne in mind when considering the modelled liabilities shown in Figure 4. First, the model corrects for the fact that drivers forget (or fail to report) some of the accidents in which they have been involved using an overall memory loss factor which is estimated from the data. This memory loss effect suggests that in a single year the number of accidents predicted by the model (and shown in Figure 4) are about 20% higher than those actually reported. Secondly, with the exception of the younger drivers, the model results are based on accidents averaged over a three year period. The young drivers in the sample reported accidents over a variable time period which ranged from a few months to several years. The effect of averaging over a number of years will have little effect on the model predictions where accident liabilities are changing slowly with time (the older drivers), but could introduce some error into the accident rates of novice drivers where the rates are changing rapidly.
Your 'standard' is how you drive alone, not how you drive during a test.