And as always with inequality issues, it's not just the disadvantaged group that is affected. Men that aren't average in all dimensions are also not represented by the dummy.
As we found out with cockpits[1], le homme moyen is a mirage. We'll see how long it takes car regulators to catch up.
But it's not just cars -- Jensen's inequality says that for any asymmetric cost function, minimising the cost of the average is different from minimising the average cost. Yet people are still -- and probably will be for a long time -- minimising the cost of the average.
There were cars back in the 70-80s that had adjustable shoulder belts that allowed you to lower the height of the belt where it's attached to the door jamb pillar behind the seat, but it's been a long time since I've seen a car with that feature.
This is just crazy: "a woman is 17 percent more likely to die ... and 73 percent more likely to be seriously injured in a vehicle crash"
Honestly, I've not paid much attention to cars for a very long time, but I have checked for that feature on the few I've owned and driven over the past 20 years and very few of them have had it.
I am sorry, but this article is drawing some sensationalist conclusions that does not directly follow.
I am not even sure what it means for a dummy to be male. Dummies are constructed to a certain shape and size, representing some average or median person. They represent a person. You may have dummies that are closer to the average man, and others that are closer to the average woman.
Now consider males who are < 5ft tall. Wouldn't they also be more likely to die? What about men who are over 6ft tall?
Pretty much anyone not represented by the dummy will have a higher risk. Gender has very little to do with this.
Some equally valid conclusions:
> People shorter than 5ft are more likely to die more in car crashes because the test dummies are taller
> People who are over/underweight are more likely to die more in car crashes because the test dummies are ...
Somehow this article seems to pull in some apparent gender bias in what is a more general bias against anyone not accurately represented by a single dummy.
I don't have all or any answers, but your comment made me daydream about a similar article talking about, say, how there should be more women in leadership positions in politics and business. And in the daydream you replied similarly, "But very tall people are underrepresented, as are the very short, the very fat, the very skinny. Pretty much anyone with a body unlike the average is underrepresented. Gender has very little to do with this."
I think the point was that you could argue your way out of seeing gender bias anywhere at all with your argument—it doesn't just apply in this crash-test situation.
And your comment makes me think that you could argue seeing gender bias in anyones response. Not trying to be belligerent by the way (I know the response sounds like I am - but really I’m seeking genuine understanding of the line being crossed in the parents logic). I do see the value in deducing gender bias even when it’s not specific as I would assume that would be a hard requirement to pass. However, at some point it’s valuable to determine that the test was not constructed in a manner to come to any conclusions with regard to gender bias.
> From 2015 Euro NCAP added a new restraint test, a full frontal impact against a rigid barrier at just over 30mph with small female dummies in the driver's seat and rear passenger seat.
and
> Improved child occupant protection (2016)
> Euro NCAP introduced two new crash test dummies representing a 6 and 10 year old child to better assess the effectiveness of restraint systems. Compatiblity with i-Size and a wider range of universal child restraints is also assessed.
> With men making up 49 per cent of the European population but 76 per cent of road deaths, using a male 50th percentile dummy allowed testers to ensure that assessments would relate to the largest proportion of accident victims.
I work for General Motors. We have test dummies of all shapes and sizes to represent children, women, babies and men. Just because NHTSA does not require them for a 5 star rating does not mean that we want women and children dying in our vehicles. We've had "smart dummies" since 2010 and have been using the data collected from those dummies to improve our models and make them more robust and safer.
21 comments
[ 2.4 ms ] story [ 55.1 ms ] threadAs we found out with cockpits[1], le homme moyen is a mirage. We'll see how long it takes car regulators to catch up.
But it's not just cars -- Jensen's inequality says that for any asymmetric cost function, minimising the cost of the average is different from minimising the average cost. Yet people are still -- and probably will be for a long time -- minimising the cost of the average.
[1]: https://www.thestar.com/news/insight/2016/01/16/when-us-air-...
This is just crazy: "a woman is 17 percent more likely to die ... and 73 percent more likely to be seriously injured in a vehicle crash"
Honestly, I've not paid much attention to cars for a very long time, but I have checked for that feature on the few I've owned and driven over the past 20 years and very few of them have had it.
I am not even sure what it means for a dummy to be male. Dummies are constructed to a certain shape and size, representing some average or median person. They represent a person. You may have dummies that are closer to the average man, and others that are closer to the average woman.
Now consider males who are < 5ft tall. Wouldn't they also be more likely to die? What about men who are over 6ft tall?
Pretty much anyone not represented by the dummy will have a higher risk. Gender has very little to do with this.
Some equally valid conclusions:
> People shorter than 5ft are more likely to die more in car crashes because the test dummies are taller
> People who are over/underweight are more likely to die more in car crashes because the test dummies are ...
Somehow this article seems to pull in some apparent gender bias in what is a more general bias against anyone not accurately represented by a single dummy.
I think the point was that you could argue your way out of seeing gender bias anywhere at all with your argument—it doesn't just apply in this crash-test situation.
> Why? All the crash test dummies are male. Even the “female” dummies the government requires in tests are just smaller versions of male dummies.
Wait, so are there or aren't there female dummies?
Secondly, does "All" refer to only the U.S., or Europe too? It seems not.
For example (see https://www.theaa.com/driving-advice/safety/euro-ncap)
> From 2015 Euro NCAP added a new restraint test, a full frontal impact against a rigid barrier at just over 30mph with small female dummies in the driver's seat and rear passenger seat.
and
> Improved child occupant protection (2016)
> Euro NCAP introduced two new crash test dummies representing a 6 and 10 year old child to better assess the effectiveness of restraint systems. Compatiblity with i-Size and a wider range of universal child restraints is also assessed.
https://www.autoexpress.co.uk/car-news/108721/women-at-great... seems to be a much better article, covering a lot of the misconceptions in the one from the topic.
Also:
> With men making up 49 per cent of the European population but 76 per cent of road deaths, using a male 50th percentile dummy allowed testers to ensure that assessments would relate to the largest proportion of accident victims.
Have you ever heard someone call women females IRL? It never sounds right.
In anything that isn't extremely formal, it reads creepy to me.
It's just like calling people men "males" - it doesn't sound right, and nobody does that in conversation.
https://www.iihs.org/topics/fatality-statistics/detail/males...
WaPo needs to hire a fact checker