US Secretary of Transportation responds to Freakonomics/GMA

Aside from the commentary about the pros and cons of carseats, which seem to vary a lot depending on whose research you believe, there's this:

"I don't know anything about science, but I do know something about data".

Here's some actual data about the "complicated and expensive" cure.

[ame="http://www.amazon.com/Evenflo-Big-Back-Booster-Seat/dp/B001H0GGMS/ref=sr_1_1?ie=UTF8&s=baby-products&qid=1263415926&sr=1-1"]Amazon.com: Evenflo Big Kid AMP No Back Booster Car Seat - Red: Baby[/ame]

[ame="http://www.amazon.com/Cosco-Juvenile-Pronto-Belt-Positioning-Booster/dp/B0012L6RJY/ref=sr_1_2?ie=UTF8&s=baby-products&qid=1263415910&sr=8-2"]Amazon.com: Cosco Juvenile Pronto Belt-Positioning Booster Car Seat in Ore: Baby[/ame]

$25-$35 is neither complicated, nor particularly expensive, and these were even highly rated by the IIHS recently. If you want cheap, there's even this one for $15 [ame="http://www.amazon.com/Cosco-Ambassador-Positioning-Backless-Booster/dp/B000UCHJA2/ref=sr_1_2?ie=UTF8&s=baby-products&qid=1263415875&sr=8-2"]Amazon.com: Cosco Ambassador Belt Positioning Backless Booster: Baby[/ame] .

And you can get them for the same price at Walmart, too, so it's not like you have to search far and wide for a great deal to get one. If a $15 backless booster is deemed complicated and expensive, we have a lot greater problems for economists to address in our society than laws governing child seat use.

He also makes a statement that makes it clear he is not familiar with scientific methods. He implies that scientists would rather use "imperfect" laboratory crash testing data than 30 years of real world crash data. The issue is not one of preference. Both can be valid methods of making comparisons. The issue is that sometimes, one or the other may be more relevant in isolating the variables you want and then making statistically and scientifically valid conclusions.

Ideally, you would prove your case using both methods. Done correctly, one correlates with the other and you have very convincing evidence. Done imporperly, you might be led to believe that a dummy in a perfectly fitted seatbelt is representative of an actual child and how they typically wear a seatbelt. Or, you might use real world evidence and omit certain outliers or subsets of data in a belief that they are not meaningful, even though they significantly change the results in favor of your hypothesis.

Laboratory tests and real world statistics can both be manipulated. Intentionally or unintentionally. This topic is a perfect example. Otherwise, there wouldn't be two sets of researchers with such credentials using the same type of testing and data, but coming to apparently conflicting conclusions.

CPSDarren said:

He also makes a statement that makes it clear he is not familiar with scientific methods. He implies that scientists would rather use "imperfect" laboratory crash testing data than 30 years of real world crash data. The issue is not one of preference. Both can be valid methods of making comparisons. The issue is that sometimes, one or the other may be more relevant in isolating the variables you want and then making statistically and scientifically valid conclusions.

Ideally, you would prove your case using both methods. Done correctly, one correlates with the other and you have very convincing evidence. Done imporperly, you might be led to believe that a dummy in a perfectly fitted seatbelt is representative of an actual child and how they typically wear a seatbelt. Or, you might use real world evidence and omit certain outliers or subsets of data in a belief that they are not meaningful, even though they significantly change the results in favor of your hypothesis.

Laboratory tests and real world statistics can both be manipulated. Intentionally or unintentionally. This topic is a perfect example. Otherwise, there wouldn't be two sets of researchers with such credentials using the same type of testing and data, but coming to apparently conflicting conclusions.

Click to expand...

I see your point, and I aggree.


Speaking of real crash data other than just numbers: Here is Volvo: They have done this for many years.
[ame="http://www.youtube.com/watch?v=5WmmyT-HNFs&feature=channel"]YouTube- 2007 Volvo S80, Accident Investigation.[/ame]

Volvo make their own seats - including a rearfacing one for toddlers and promote/sell some seats from car seat manufactorers

Lena

CPSDarren said:

He also makes a statement that makes it clear he is not familiar with scientific methods. He implies that scientists would rather use "imperfect" laboratory crash testing data than 30 years of real world crash data. The issue is not one of preference. Both can be valid methods of making comparisons. The issue is that sometimes, one or the other may be more relevant in isolating the variables you want and then making statistically and scientifically valid conclusions.

Ideally, you would prove your case using both methods. Done correctly, one correlates with the other and you have very convincing evidence. Done imporperly, you might be led to believe that a dummy in a perfectly fitted seatbelt is representative of an actual child and how they typically wear a seatbelt. Or, you might use real world evidence and omit certain outliers or subsets of data in a belief that they are not meaningful, even though they significantly change the results in favor of your hypothesis.

Laboratory tests and real world statistics can both be manipulated. Intentionally or unintentionally. This topic is a perfect example. Otherwise, there wouldn't be two sets of researchers with such credentials using the same type of testing and data, but coming to apparently conflicting conclusions.

Click to expand...

And this is presicely the reason why I don't understand why people can't get that rear facing is significantly safer than forward facing. The sled tests are there. The real world tests are there. The data investigation of real crashes is there. What is it going to take to get the message out? And to get people to understand it?

And Lena, I can't believe that the US wouldn't allow volvo to market their seats to their owners, or that Volvo wouldn't redesign some aspects of their seats to suit the US government. I feel they both failed us Americans in some way, kwim?

And Lena, I can't believe that the US wouldn't allow volvo to market their seats to their owners, or that Volvo wouldn't redesign some aspects of their seats to suit the US government. I feel they both failed us Americans in some way, kwim?

Click to expand...

They are not universial, thereby I have said that they lack the LATCH system. They would need to be redesigned from the ground up in order that LATCH can be added to them. There are also the front braces and the floor support legs that need to be added to the USA standard for car seats. the journey would be too long and too exspensive I guess.

Lena

It was more of a rhetorical, I can't believe they couldn't release them statement. I know "why" they wouldn't, I just think it is stupid