Another stats blunder
Stats Goofs | 10:22 am Saturday, Dec 17 2005 |
They’re on a roll today at The Age.
From ‘Deck the gums, it’s Yuletide Down Under’:
BOOZE
Almost 930 drivers were breath-tested between December 23 last year and January 6, with the highest reading recorded .307. Of those tested, 184 recorded readings of .05 to .069 ($314 fine and 10 demerit points), 248 drivers registering .07 to .099 ($314 fine and six months licence cancellation) and 195 people recording readings between .100 and .149 ($440 fine and between 10 and 14 months licence cancellation).
A further 77 drivers with blood alcohol readings of between .150 and .307 had their licence immediately suspended and were given a summons to appear in the magistrates court.
Hang on, are you saying 704 out of 930 drivers that were breath-tested were over the limit?
Over 75%? Surely not!
My guess is that 930 of the thousands of tested drivers were given a second breath-test in the booze bus or down at the station, and of those 930, a total of 704 drivers were over the legal limit. That leaves the remaining 25% at 0.04 or below — meaning they were close enough to the limit for a second test, but not doing anything illegal.
(It still doesn’t account for any drivers on a zero blood alcohol limit who may have been at 0.04 below and still over their prescribed limit).
But surely this figure of 930 over the 2-week period does NOT represent the total number of drivers tested.
According to a media release on the Victoria Police website (Road Safety Blitz in Stonnington),
- 1,700 drivers were tested in just one night by two Booze Buses
- Just five of those drivers were charged for exceeding the prescribed blood alcohol limit
Let’s see — five out of 1,700 exceeded the limit … that’s less than 0.3% of all drivers.
So the article is wrong on two counts:
- An actual “strike rate” of 0.3% versus 75% — it shows how incredibly far off the mark the above claim is.
- 930 drivers tested during the festive season, versus 1,700 in just one night in one location — again, incredibly misleading.
Yet another example of media errors that have made it through all writing and editing stages to publication.