Consumer Reports Flubs Child Seat Study

The magazine that likes to be known for pointing out dangerous or defective products had to do a bit of spin control itself today. Earlier this month, Consumer Reports released a study that showed 10 out of 12 popular child-safety seats on the market failed its crash tests.

Turns out, that’s because the magazine used a significantly tougher test than the federal government does to come to its conclusions.

In testing child-safety seats, the National Highway Traffic Safety Administration only performs frontal crash tests simulating speeds of 30 mph; it forgoes the use of side-impact crash data, which some see as a major flaw because today’s new cars are also given side-impact crash-test ratings. That’s why Consumer Reports wanted to step in and do its own side-impact tests.

The government’s side-impact tests for new cars simulate a crash at a speed of 38 mph. Consumer Reports intended to apply the same standard — 38 mph — to show what such a crash would do to popular child-safety seats. Instead, it carried out the tests at more than 70 mph, or highway speeds. Obviously, that type of impact is not typical. After Consumer Reports released its study, NHTSA retested the seats in side-impact crashes using its regular 38-mph procedure and they all passed and did not become loose.

An official announcement from Consumer Reports about the erroneous study came today; the magazine apologized for the error but didn’t mention how drastically off the tests were.

Whatever the result, we can say this: The two seats that passed the 70-mph test will probably still sell better than the rest.

[Consumer Reports Recalls Car Seat Study, CNNMoney.com]

Comments

It's also worth noting that CR outsourced the testing to a vendor. At least that's what I saw on MSNBC last night.

Just goes to show you, if you want something done right, you gotta do it yourself.

While I'm on the topic of CR, I gotta say, I've never really trusted their product "opinions," especially in the car space. They seem as biased as anybody else. But I've tended to hold their quantitative analysis in pretty high regard, whether it was about cars or toilets or the viscosity of peanut butter.

The news (to me, anyway) that they're outsourcing some of that analysis makes me a lot more skeptical. Companies use vendors either because the vendor can do it better, or because the vendor can do it cheaper. Usually the latter.

One way vendors do it cheaper is by cutting corners. Maybe they hire less-qualified staff. Maybe they hire programmers in East Timor. Maybe they just don't do the QA they should.

Every wonder why almost NOBODY ever fails a drug test when they're applying for a new job? It ain't because golden root seal or herbal tea actually works -- it's because the drug-testing vendors don't really test a lot of the samples! They just send them back as "pass."

That's my theory, anyway, and I'm sticking to it.

However, according to another article (http://www.childseatcenter.com/article4.html), Consumer Report also recommended a car seat that was later recall!