Blessed by the Potato

I’ve been saddened by a rash of product shutdowns by Google. Lately the announcement that Reader was being shut down is making waves, but that’s the third Google product I use that got shut down.

First there was Sync, which was awesome. It was a blackberry app that synced my contacts, some other crap, and importantly, my calendars with my BB. After shutting it off there is still a mechanism to sync the blackberry calendar and contacts list with Google’s — but it is far inferior. In particular, Google Sync was able to figure out that I can have more than one calendar in my account, whereas the crap I’m left with will only sync the main calendar, forcing me to manually sync up with my shared calendars.

Then iGoogle was given the kiss of death. It still operates, but constantly reminds me that it will be “sunsetting” soon. For those that didn’t try it, it was just a way to customize your Google homepage. You’d have your usual search box, and then boxes for RSS feeds or little widgets. It was basically what the Windows 8 start screen was except better and properly encapsulated in a web browser. And now, Reader will be going too.

The thing is, I don’t understand why they’re being killed off. Was Sync really so hard to maintain? Does syncing all my calendars through an app take up much more Google bandwidth than having my BB sync one?

We bought the Graco Snugride 35 carseat initially because all the safety research suggests that keeping kids rear-facing longer is safer. Rated up to 35 lbs, we figured that this car seat — though large and heavy — would keep Blueberry rear-facing until she was pretty much two years old. Though it’s large for an infant carrier, I was able to fit it in the Prius and still manage to get my seat to a decent position for driving (it’s about an inch further forward than I had it when I positioned my seat without any such constraints — not the most comfortable position but a workable compromise). Blueberry is very tall for her age (obviously a mix-up at the hospital), and though she still has over 10 lbs to go before hitting the weight limit she’s getting close to the maximum height for her infant seat. Time to move up.

So now we’re off shopping for convertible car seats, the next step up that can be either rear- or forward-facing. With these larger seats, it’s almost impossible to find ones that can fit behind a front seat well enough for me to drive or for Wayfare to comfortably sit. I’ve been checking various forums for tips and reviews and pictures of how they fit, and it seems like the two on our shortlist are the Britax Marathon/Boulevard or the Diono Radian. I’ll spare you my pro/con lists, coin-flipping, and hand-wringing on this decision (though feedback on those seats is welcome in the comments).

What really got me in our search was the oft-stated fact that carseats are improperly installed some huge portion of the time. I heard numbers ranging from 80% to 95% depending on the source, and it got me thinking: where does this bit of conventional wisdom come from? I’ll grant that installing the old-fashioned way with a seatbelt is difficult both in terms of skill and strength required, but I really had no issues with the LATCH install. After all, that’s what LATCH is supposed to help with. Plus, the epidemiology data all says that kids in car seats are safer, so either the install error-rate is over-stated, far more people are managing to get/pay for a professional installation, or seats are safe even if installed incorrectly. I started to wonder just how true this conventional wisdom was, or if perhaps this factiod had been invented by the stores offering a $25 installation service and picked up by the media, so I went off in search of a source.

There are some NHTSA reports that seem to be the origin of these figures. This one, for instance, gives a high error rate for installation, topping 95% for first-time installers, who in this study (or a similar one I just read) were recruited from a university’s volunteer pool (i.e.: first-year psych students giving their very minimum effort for $10 and a course credit).

The most common error is loose installation: a carseat, when properly installed, is supposed to be able to move less than an inch. Now, a carseat that can be wiggled an inch and a half is not meaningfully more dangerous than one that can only be wiggled an inch; likewise, the carrying handle for a removable carseat has a specified position for use in the car for each brand (and it is often different for each model) — though many first-time installers got it wrong, it’s also not usually critical. If they apply a severity score, then “only” about 30% of seats were incorrectly installed in a really bad way. The good news: the error rate drops in half once parents/caregivers who have carseat experience are tested, rather than novices. The bad news: that’s still a nearly 50% error rate. To pick out one more interesting factoid, there was a higher error rate for those who drove cars with leather seats.

I’m surprised that even digging into the data, the “legitimate” error rate still appears to be shockingly double-digits high. That really says that something needs to be done to make carseats easier to install safely. Some kind of standardization is most likely the answer: either continue with LATCH but standardize the connectors, or create a universal base that the manufacturer’s individual seats can clip into. Angle adjusters with a wide range of motion are also likely going to be needed — far too many official installation instructions include the use of towels or pool noodles (sold separately) to prop up one part of the base, which is frankly ridiculous. Many require a great deal of strength to tighten properly, or that the adult put their full weight on the seat to jam it down into position — a ratcheting belt-tightener would be a great feature on many of these seats.

As an aside (and not necessarily a product recommendation) this car seat is a neat one from a human factors point of view, with sensors and a display to help ensure correct installation. The video there is only about a minute long if you want to go have a watch.