Ridiculous statistics

Published: Wednesday, 07 May 2014

As we all know statistics can be used selectively to justify almost any statement .... and are frequently so employed to substantiate statements that cannot otherwise be given credibility, writes Jimmy Lockwood.

I well remember a speech many years ago by a senior officer of a northern police force quoting the following statistics and drawing possible conclusions.

"Let's look at accident statistics. In 30% of all reported road accidents alcohol played an import factor and in those cases the driver was found to be over the limit. However, the statistics also show that in over 60% of accidents alcohol was not a factor. Most accidents occur when the driver is sober.

Clearly therefore you are much less likely to be involved in an accident if you have had a drink!"

The Environment Agency

I guess the EA would support that statement as being the correct interpretation of the data! (Victor: You must be joking) Most of us, I suggest, are reasonable beings and would not agree that drinking makes driving safer.

What is the EA trying to prove with its statement?

What are the sources for the base data? Are they providing any breakdown of the types of accident? The times of day? The location? The causes? Are those involved boaters— using the lock for its intended purpose or just towpath walkers?

I suggest that any statistics used, without substantive data, to justify bald statements should be treated with the contempt they deserve and dumped in the nearest litter bin.