From Paul Fox
In your remarks on online surveys you claim “we vote alternate ways to avoid skewing the results” (Feedback, 22 February). Might I point out that you will skew the results, in favour of the minority position. Suppose, for example, that in an admittedly poor response to such a survey, 2 vote one way and 1 the other. Feedback then adds a vote for each side and a 2 to 1 majority slips back to 3 to 2.
I trust that you will be suitably horrified to discover that you have been, albeit inadvertently, corrupting statistical data in this way.
Stoke-on-Trent, Staffordshire, UK
