Retorts

Lies, Damned Lies

I found Miles Maxted's [Oct] article on New Zealand polling techniques intriguing, if not a little disturbing. It was once the practice to cease announcing poll results in the week leading up to the election. Some of the early surveyors were concerned at the potential power their polls had to affect or change people's voting behaviour. It certainly made me feel as if things were hopeless...

One area of bias the article didn't explore was the interpretation of the data once it's been collected. If the actual statistics theoretically underpinning everything are as bad as made out, then the interpretations have got to be really rotten. Doesn't leave you with a lot of confidence in what They say we think.

Eleanor Bruce, Auckland