Wrong end of the magnet

Just a hint? When all the evidence you provide says "no," your hed shouldn't pose a question. (And your story shouldn't say "yes," but at that point, we're getting perilously close to killing the story altogether.*)

The tornado that tore through Raleigh last weekend was a reminder that twisters often strike major metropolitan areas, despite a longstanding myth that the storms are a mostly rural phenomenon.

But you don't have to look to Raleigh for an example. Statistics show Charlotte and Mecklenburg County are a tornado magnet, of sorts.

Oh, where to start? Well, man-bites-dogwise, if tornadoes "often" struck major metropolitan areas, they wouldn't be news, would they? I think we're trying to say that tornadoes aren't necessarily less frequent in cities than in the sticks, but "not less frequent" and "often" aren't the same thing. Two, no; you don't have to "look to Raleigh for an example," but since several of the storm system's deaths were in Raleigh, this seems like an inappropriate time to be jumping around and demanding that the world pay attention to Little Old Charlotte. Third, the hed says "Mecklenburg" (for you outlanders, the county that houses Charlotte), which is a bit different from "Charlotte and Mecklenburg."

And fourth, they do not. The statistics don't show any such thing, as the story and its accompanying infobox make abundantly clear.

Statistics show that Mecklenburg has been hit by more tornadoes than any other county but one in the region during the past 60 years.

Scientists say the reason is a small part science and a lot of common sense.

I think that's "reporter says," not "scientists say." Part of the problem is standard journalistic carelessness with what's being measured. The topic here -- unless I'm badly misreading things -- is tornadoes that are reported. That's directly related to the common sense part, which seems to be that if a tree falls in the forest and no one sees it, it (surprise!) isn't seen:

"Where you have more people, you have a greater chance of someone seeing the tornado and reporting it," he says. "Look to your more populated areas, and you'll probably see more reports."

In addition, he says, people are more likely to notice a tornado's damage in a developed area than, say, in a county that is 70 percent wooded.

Which could explain why the county that holds the biggest city in either Carolina has almost as many reported tornadoes as a much less densely populated county across the border -- and only one more than two other counties in the "region" (which often means the neighboring counties in North Carolina and three in South Carolina). Starting to sound a little less magnetic now?

Part of the whole "objectivity" thing is testing your observations, meaning (among other things) that you ask appropriate questions when you have the "scientist" on the other end of the line. In cases like this, it's always good to raise the same question you ask of all "poll says" and "study says" stories: How likely is it that these apparent differences -- 19, 20, and 22 tornadoes in 60 years -- came about by chance? If the answer is "pretty likely," then your story is "No tornado magnets here."

To make matters worse, you could consult the handy infobox, which puts Charlotte and Mecklenburg last among "major regional cities": Columbia, S.C. (Richland County), 34; Raleigh (Wake), 30; and Atlanta (Fulton), 22. If Mecklenburg County is a "tornado magnet," somebody must be pointing the wrong end at the tornadoes.

Not every story has a local angle, and not every Stupid Question has to be answered. The main statistical point to take away here is familiar: There's one born every minute.


* Not that such an outcome would pose a major threat to the well-being of the free world or anything.