> > Do a search for "SUV fatality rate," or something similar. There's huge amounts of data available.

> You simply cannot base an argument on one article found on the Internet that may or may not have a slant to it.

I read the first score or so such articles Google pulled up, including all the ones you listed. I linked to the one I did because it was the only one with the specific stat I was looking for, which was deaths per miles driven, sorted by different vehicle types. Sam mentioned many different ways the stats can be phrased for slightly different purposes. Since I was looking for only that specific stat, I did not feel it necessary to find that stat in multiple places. And the reason I was looking for that specific stat was because it is the only one that makes sense to me in a discussion about safety. The stat had to be about fatalities, not accidents, because accidents by themselves are no indicator of safety. It had to count fatalities from all accidents of all accident types proportionally, so as to give an honest guideline to real safety. It also had to be on a percentage basis of the vehicle's use, thus the 'miles driven' requirement; otherwise, the most popular vehicle would have the most accidents and the second most popular the second most accidents, and so on. If anyone thinks of a more apt statistic, that would always be welcome.

>Furthermore, you can't cite one type of product (in this case, the Ford Explorer) and use that as your evidence.

The article I quoted had one type of product (the Ford Explorer) and its fatality rate, but it also had the fatality rates for *all* SUVs and *all* cars, which was, of course, the salient point. I chose not to edit all of the sentences in order to remove references to the irrelevant Ford Explorer.

>I assume your family owns a Ford Explorer?

No, nor a Ford, nor an SUV, nor anything larger than a Tercel. While bias is often easy to dismiss with more bias, the numbers given remain the same until a competing set of data shows otherwise. I'm not emotionally invested in this discussion at all. It just bugs me when people make claims that their statistics don't actually support, such as the 'higher rollover rate = less safe' idea mentioned earlier.

> Ell"NEVER use one source!!!"myruh

Agreed. But neither should one dismiss a source for being only one source.