The fall. And winter is the safest.

Tuesday July 14, 2009
Surprise: Fall most deadly driving, winter safest

DAVID N. GOODMAN
Associated Press Writer

DETROIT (AP) - Icy winter roads and vacation-clogged summer highways might seem the most dangerous for motorists, but new research says fall is when driving is at its deadliest.

A new research report has found October has the year's highest death rate per distance driven. In fact, the danger of dying in a crash is 16 percent greater in October than in wintry March - the safest month of the year.

The findings by Michael Sivak at the University of Michigan Transportation Safety Institute in Ann Arbor are reported in the July issue of the journal Traffic Injury Prevention.

Sivak analyzed monthly crash death figures for the U.S. from 1994 to 2006. He says October had a death rate of 10.2 per billion kilometers, compared with March's 8.8 per billion.

 

I hate mishmash statistics. Why can't they post the actual number of people killed, instead of dividing it by miles driven? How about something meaningful like deaths per day. Then we could see if a particular month is truly "deadlier."

Of course the average based on miles driven will be shorter in summer. People drive considerably farther distances in shorter periods of time. It's called vacation.

 

detroit article.. statistics can be misleading here in my homestate.

 

The problem with a flat number is it hides the significance. For example, 10.2 deaths per billion kilometers is the same death rate, regardless of where it is. But if you say there were "10 deaths last year", it means something drastically different for Los Angeles CA than it does for Nome Alaska. LA, no one notices, Nome, that is half their population.

The biggest thing to remember from any 'statistics' like this though is:
"What information reveals is important, but what it conceals is vital"
or
"Do not consider what information says if you are not also going to consider what it does not say"

What this stat doesn't say is why, or the crashes that almost kill you, but only leave you horribly mangled instead. Just like the "speed is a factor in a 3rd of crashes" isn't lying, but it is putting incorrect focus on one of several factors, when the other factors (such as intoxication, distracted, inexperienced, faulty equipment) have a higher 'involved' rate.

 

Presumably wet leaves and/or black ice, and low angle sunlight in the Fall when otherwise no special driving care seems necessary?

 

Could be. Here is a longer article that says this:

As for complaints about the methodolgy, it says right up front that the study is measuring "the year's highest death rate per distance driven." Of course there might be more deaths in months with more driving. Significantly, there ought to be fewer deaths in months with less driving. October has the distinction of the most deaths per distance driven.

That's an entirely valid analysis. If you were studying homicide rates, it would make no sense to compare the total number of deaths in Los Angeles with the total number in Des Moines or Peoria. What you would want is the per capita rate -- the number per 100,000 people.

 

I like how they use March for winter instead of say... January. In some states March is already summer... Fools.

Hell we hit 90's in March in NY, occasionally

 

"In theory, astronomically, the equinoxes ought to be the middle of the respective seasons, but temperature lag (caused by the thermal latency of the ground and sea) means that seasons appear later than dates calculated from a purely astronomical perspective. The actual lag varies with region, so some cultures regard the autumnal equinox as "mid-autumn" whilst other treat it as the start of autumn."