31.01.2015 Views

Glossary of Data Variables for Fatal and accident causation ... - ERSO

Glossary of Data Variables for Fatal and accident causation ... - ERSO

Glossary of Data Variables for Fatal and accident causation ... - ERSO

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Manual For SafetyNet <strong>accident</strong> <strong>causation</strong> system (SNACS)<br />

km/h. From this you can say that the speed is more than adequate (given the<br />

speed limits), but the stretch <strong>of</strong> road with an unblocked view <strong>of</strong> oncoming traffic<br />

is too short. The latter part would indicate that the distance available <strong>for</strong> the<br />

overtake is too short, making Distance a suitable genotype. However, a part <strong>of</strong><br />

normal driver education is to teach the driver not to overtake unless there is a<br />

sufficient stretch <strong>of</strong> road with a free view, <strong>and</strong> hill crests are particularly<br />

inadequate <strong>for</strong> overtaking. Given this, the phenotype that best describes the<br />

situation would be Timing: premature action. The overtaking should have been<br />

per<strong>for</strong>med later on at a more appropriate stretch <strong>of</strong> road.<br />

In practice, this does not have to be a very complicated problem, since the<br />

linking table is constructed in such a way that phenotypes which are close also<br />

have quite similar linking from the genotypes. This means that the choice <strong>of</strong><br />

phenotype does not have to be perfect, you can still get to the causes you need<br />

<strong>for</strong> the analysis.<br />

6.5.4 Subjective judgement<br />

As the above reasoning shows, the SNACS methodology clearly calls <strong>for</strong> the<br />

investigator to reason from a “normative” st<strong>and</strong>point, i.e. the investigator has to<br />

have an opinion on what would be normal or reasonable behaviour <strong>for</strong> a drivervehicle<br />

system in the situation at h<strong>and</strong>, because otherwise it will be difficult to<br />

classify what the deviation from the norm is. Also, the reasoning shows that the<br />

choice <strong>of</strong> phenotype will be partially based on the investigator’s judgement. In<br />

both these regards, a SNACS analysis will very clearly always be “subjective” in<br />

some sense <strong>of</strong> the word.<br />

This can be considered both an advantage <strong>and</strong> a disadvantage. People who<br />

consider it a disadvantage will say that <strong>for</strong> a SNACS analysis we can never be<br />

sure that two investigators will per<strong>for</strong>m the analysis in the same way, since<br />

subjective judgment is involved. This would be quite devastating <strong>for</strong> the<br />

methodology, were it not <strong>for</strong> the fact that all methodologies which analyse<br />

causes suffer from the same problem. As David Hume elegantly showed in his<br />

treatise on human reason, causes typically cannot be observed; they have to be<br />

inferred by reasoning. This means subjective judgement will always be involved<br />

when analysing causes, no matter whether you are investigating injury<br />

<strong>causation</strong> mechanisms in passive safety or <strong>accident</strong> contributing factors in<br />

active safety.<br />

In light <strong>of</strong> this, we find it preferable to have a methodology which transparently<br />

shows that human judgement is involved, instead <strong>of</strong> burying that fact under<br />

protocols designed to look “objective”. As <strong>for</strong> the problem <strong>of</strong> different<br />

investigators per<strong>for</strong>ming different analysis, the best we can do is to clearly write<br />

down the reasons we have <strong>for</strong> making our choices <strong>and</strong> then discuss these<br />

reasons within the larger group <strong>of</strong> both your own <strong>accident</strong> investigation team<br />

<strong>and</strong> several different teams meeting <strong>for</strong> case review. As has been pointed out in<br />

the philosophy <strong>of</strong> science many times; over time a consensus within a larger<br />

group or community is as close to the truth as any methodology analysing<br />

causes will ever be.<br />

Page 128 <strong>of</strong> 215

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!