http://www.washingtonpost.com/wp-dyn/content/article/2011/03/13/AR2011031303926.html
There are really no federal regulations for seats that go to or past 65 lbs? These are somewhat new. Are they being thrown at the market without real (reliable and trustworthy) testing?
There are many federal regulations for higher weight seats. Just not ALL of the same ones that exist at lower weights. It's worth noting that if a child restraint also fits a smaller child, then that restraint is held to the standards applicable to the smaller child. They test with all the dummies that fit, not just the biggest one.
And there is no dummy that is higher-weight? Nothing reliable? Read this article.
None of the child dummies are outstanding representatives of actual children, for many reasons. One is that the only way to learn how a child's body reacts under crash forces is to take a child's body and intentionally subject it to crash forces. Clearly we can't do that with living children. The alternative (and how adult dummies were developed) is to take cadavers (bodies donated to science) and "crash test" them. Nobody really wants to do that with children because (1) parents don't want their deceased children's bodies treated that way, and (2) the scientists don't really want to go there, either.
Children aren't small adults. Their bodies are not tiny adult bodies. But for now, that's what we have for crash test dummies.
There is a newborn dummy, a 12 month old, a 3 year old, a 6 year old, and a 10 year old. The 10 year old dummy is not used in testing child restraints because of the lack of biofidelity (it doesn't act like a 10 year old child's body, not even close). IIRC, the head snapped down to the chest too hard, giving high readings for head injury, even when a real child's head wouldn't move that way and therefore wouldn't have the head injury. The choice was to use the dummy and have bad data for the head injury, or position the head in a funny way and end up with bad data for all the other injury criteria. They chose neither, instead adding weights to the 6yo dummy's spine.
How on earth do these seats make it to market? How can I assume ANY are safe? The article is scary. Why will it take until 2013 to get a dummy developed?
I don't want my higher-weight harnessed kid to BE the dummy until then.
The seats make it to market because they pass all federal standards for child restraints.
I touched on dummy development already.
Your choice is to have your child BE the dummy, or don't go in a car. It doesn't matter if your child is in a high weight harness or a belt positioning booster. We just don't have the dummies to test these seats.
We DO know that a child is better protected in a properly used harness or booster than in a seat belt alone. Accident statistics clearly show that child restraints do their job.
If you want to get worked up over something, take a look at the crash test requirements for the back seat passengers in new vehicles. I'll give you a hint, there aren't any. Of course, that doesn't mean that the vehicle manufacturers never put a dummy in the back when they are doing product development, but they are not required to do so, and there are no standards. That doesn't mean that back seats are automatically unsafe; in fact, crash statistics clearly show that the back seat is safer than the front seat.
And why would there be no side testing at all? These can be the most deadly. I don't get why the industry is lagging so much.
They can't use the same dummies for front and side testing. The sensors which are designed to detect front to back movement won't work side to side. Also, the types of injuries sustained in front and side impacts are different. They have to develop a whole new set of dummies. Which brings us back to the conundrum about how to go about doing that.
Child restraint standards have only existed for 30 years. I believe 1981 was the first year with any standards whatsoever for a child restraint. Adult vehicle occupant protection has existed far longer; seat belts have been required in vehicles since 1965, and existed as an option before that.