It’s a better car, except when it isn’t

There are times when I just can’t figure out Consumer Reports.

In November, there’s a sidebar in the Cars section that says the following:

We now recommend the [Chevrolet] Volt plug-in hybrid after new data from our 2011 Annual Auto Survey shows it earned much better than average reliability. Very few of the 116 Volt respondents had any serious problems in the first few months of ownership.

Which seems reasonable to me. All the major hybrids — Toyota, Honda, Ford — are showing better-than-decent reliability figures, perhaps because of the extra development time that goes into hybrid design: you’ve got to have pretty tight tolerances, or it won’t work at all. If the sample size seems small, well, there are only a couple of thousand Volts out there; it’s at least as statistically valid as responses on, say, 15,000 Camrys. (If you own a Porsche, your mileage may vary.)

In the CR road tests, the Volt scored an okay, if not inspiring, 67, about four points behind the baby Lexus (CT200h) hybrid.

None of this would pose a problem except that in the same issue, they test a Hyundai Sonata Hybrid, their sample of which proved to be deeply flawed: it scored, they said, “too low to recommend.” The Sonata rolled up a score of 69, two points above the Volt.

Now it was my understanding that CR’s reliability ratings and road-test scores had nothing to do with one another. The criteria for Recommended:

“… did well in our road tests, had average or better reliability in our subscriber survey, and performed at least adequately if included in government or insurance-industry safety tests.”

The safety details for the Sonata Hybrid, as given, look fine to me, and better than anything else tested in that issue. Something here doesn’t add up.





1 comment

  1. Dan B »

    5 October 2011 · 4:32 pm

    This is my understanding, subject to human error:

    The road tests are based on one or a few cars that CR staff gets to actually drive themselves. The reliability ratings are based on the feedback from the readership. A car can score well on the road test, but if something consistently breaks at 12K miles, the readership will let them know. Or the readership responses might be so “all over the map” that they are not statistically-reliable.

RSS feed for comments on this post