Saturday, November 6, 2010

This new BC Angus Reid Poll

Getting an answer that can be quantified and function as a useful measure of public opinion is hard to do at the best of times.   Many of the questions pollsters ask do not account for what people mean when they answer the question.    One of the questions I hate is the one when they ask if their opinion got better or worsened about a politician.   These numbers are utterly dubious because some that hates a politician is still very likely to say their opinion worsened even if if has not changed.  

With all that in mind, I have some major problems with the new Angus Reid poll and what the numbers show about the BC political scene.   First off all, there is no indication how they have accounted for non-voters, roughly half the adult population.  Until a pollster can explain why they are wrong on this number but should have any of their data taken seriously, the data they produce is not functionally valid as a measure of public opinion.

This poll is being touted because part of the poll indicates that the public has reservations about the NDP.   First off, the question was who would you like to form the government in 2013, not who would you vote for.   Second, this sample would be the whole sample of people as this is not a question about an individuals intentions.   Someone that is not voting can have an opinion on who they would like to see form the next government.

The answer is also being misrepresented in the media as a measure of popular support for the parties which it is not.   The two numbers for the Liberals should be combined to give a cumulative preference for a Liberal government.   If one believes these numbers, 34% want the Liberals to win 2013, 32% the NDP and 34% someone else or not sure.    But what do those numbers really mean?   At best that we have a divided province where 1/3 wants the NDP, 1/3 wants the free enterprise coalition and 1/3 does not really care   Real rocket science there, all it is saying is that the people of BC are still here and alive and hold the same broad views as to who they want to have as government as they have had since 1940.

The real impact of the poll is the changes it pushes.   It clearly is one that is going to be used by people within the NDP to get rid of Carole James.   As much as I think the NDP would be better off without Carole James as leader, it is wrong to use information gleaned from a poll that can not show that it is an accurate reflection of public opinion or what the answers mean.    The question about who people want to govern gives us nothing new to work from and is not a negative comment on the NDP.

No comments: