Media Watch: Good Graphics but No Cigar

[Don Coombs <dcoombs@uidaho.edu>]

Some newspapers have problems with numbers. As noted on Vision2020
earlier, the Spokesman Review has dedicated lots of space to the "killer
highway," U.S. 95. This is a popular piece of journalism because almost
all of us drive, and who wouldn't want to drive on a safer highway?

The Spokesman Review deserves good marks for its graphics, but:

Is Highway 95 really a killer highway? To be more specific, is driving on
Highway 95 in Idaho a lot more dangerous than driving on the average
non-freeway U.S. highway? Or on the average Idaho highway? Could be, but
you wouldn't know it -- the numbers weren't there in the Sunday piece. 

What is needed is a comparision of accident rates (adjusted for length of
highway and traffic density) for Highway 95 and other mostly two-lane
highways in Idaho or Washington or Oregon or Montana. (Just ONE comparison
would be better than none. And, as I found out for myself, those numbers
are easily available.) (The Spokesman Review said "Highway 95 consistently
has a higher accident rate" than similar Idaho roads, but the
paper provided no numbers.)

The Tuesday (Jan. 25) coverage in the Spokesman Review said "UI
researchers link crashes to road width" but there was no indication how
strong the link was. (Look at news stories about the link between smoking
and lung cancer, or being overweight and having a heart attack: It's
standard practice to report in one way or another how strong the link is.)

This morning (Wednesday, Jan. 26) the Spokesman Review included some
numbers in its story. It identified the most dangerous stretches and
said: "The worst spot, on the southern boundary of the Coeur d'Alene
Indian Reservation, has a startling two accidents for every 1,000 cars
that pass it."

That's straightforward writing, but it's wrong -- so wrong that it should
have prompted someone at the newspaper to check. I found out that the
average traffic load between Moscow and Coeur d'Alene is 3,510 vehicles
per day. So you do the math: That's 3.5 (thousands) times 2 (accidents per
1,000 cars) which gives you:

7 accidents at that "spot" on the average day.

It's not possible for there to have been 7 accidents per average day on
the entire Moscow to Coeur d'Alene stretch of highway. That's because from
Coeur d'Alene to the Latah County line (a somewhat shorter stretch, I
admit) there were only 116 accidents in 1998! That's less than one
accident every three days. (And that includes non-injury accidents where
there was just property damage.)

One more item of interest: My casual telephone call to the Idaho Office of
Highway Safety in Boise revealed that Highway 93 (from Jackpot, Nev., to
Arco, Challis and Salmon) has had a higher accident rate the last four 
years than Highway 95. (This is taking into account different
lengths of highway and different traffic density, and is true whether
you are counting collisions or just fatalities.) I guess Highway 93
doesn't go close to a big newspaper with a graphics department.

Newspapers often don't do well with numbers, perhaps because the young
people who end up in journalism are better at writing than at
calculating. But some stories have to have numbers (meaningful, correct
numbers) in them or the reader isn't well served.

Sources of information in this posting: The Idaho Transportation
Dept. and its Office of Highway Safety.

Don Coombs
208-882-8720