 |
| Actually we do: 45% were in favour and 41% opposed |
Just over a month ago I
wrote about the North Wales Police Authority's shameful misrepresentation of the results of an statistically invalid internet survey which they conducted together with the South Wales Police Authority to supposedly gauge the Welsh public's opinion on elected police commissioners. Despite only 884 people taking part (and only 82 of them coming from North Wales) the NWPA then released their interim findings to the Daily Post, stating categorically that residents of North Wales didn't want a elected police commissioner -- except that they
actually did, with 45 percent either in favour or somewhat in favour, compared to 41 percent who were against. In so doing the NWPA were clearly guilty of wilfully misrepresenting the results of a sham public consultation in order to oppose the new policy and protect both the status quo and their own sinecures.
In response a regular commenter to this blog then contacted both the NWPA and SWPA by letter asking them to comment on why they had released the results of a survey which breaks so many elements of good practise on polling, namely:
"(1) The sample is not representative of the population because it is self-selecting. In other words, only those who feel they have an opinion on the topic will trouble themselves to respond. There is some background information gathered about respondents, but this is not complete, and crucially lacks information on socio-economic background. There is therefore no robust means of correcting for the sample that you do gather.
(2) The poll asks for the respondents' sex, but the responses are not broken down by sex. In contrast, and perhaps in a nod to political correctness, the fraction of 'ethnic minority' backgrounds is quoted (but not how they 'vote' on the poll).
(3) The response rate per head of population is exceptionally low; Anglesey (population ca. 70,000), has only 7 respondents - 1 in 10,000. Even as a total, 884 responses is well below the accepted minimum representative (which it anyway is not) poll size of about 1000 respondents.
(4) You do not provide the margin of error (the confidence intervals) for the poll. If it were quoted, the confidence in the results would appear to be low.
(5) There may be bias in the questions asked, and so stray into a 'push-poll'. It is not clear who drew up the poll, but it is clear why it is was drawn. From the general content of the report so far, there seems to be a bias against the concept of elected commissioners within the NWPA."
The response almost one month later?
NWPA: no acknowledgement, no response.
SWPA: acknowledgement received, no further response.
So there you are: you can choose between (a) a number of unelected appointees/quangocrats who clearly do not feel they have to respond to reasonable questions from members of the general public; or (b) an elected commissioner who will be held directly accountable to residents for police priorities and performance.
I know which option I would choose.
On a side note, the acknowledgement from the SWPA said that the letter would be passed on to the Chief Constable -- which begs the question of just how independent the Police Authorities really are. Why would the chief constable of South Wales need to be informed about a complaint regarding the methodology of an internet poll carried out by a body supposedly at arms length from the Police?