Archive for July, 2018

Deal or no deal Brexit, support for the far right and the death penalty – Polling Matters

Posted in Politics, Polling Matters on July 26th, 2018 by Leo – Be the first to comment

On this week’s Polling Matters podcast Keiran and I look at the levels of public support for a no-deal Brexit and how that may factor in to decisions in Westminster.

Also on the podcast, we look at polling on support for the far-right and other potential new parties by YouGov as well as public support for the death penalty in light of Sajid Javid’s decision not to seek assurances that suspected British Jihadis will not face execution when facing trial in the United States.

 

ComRes and We, The People back with another loaded poll

Posted in Bad polling on July 16th, 2018 by Leo – 1 Comment

ComRes and We, The People, a new secretly-funded right-wing lobby group, are back with another loaded poll – a few weeks after they got the Mail to cover their pro-Brexit poll.

This poll got the Mail’s front page today, with the claim most people think the police have lost control of the streets.

As with the pro-Brexit poll, the design of this survey is likely to have helped the We, The People get the answers they wanted. Here are a few of the problems with the poll:

  • An early question asks whether respondents have seen a police officer on their street in the last year – a ridiculously high bar (how much time do you actually spend looking at your street? A couple of minutes a day?). Not surprisingly, most people said no, preparing them to think policing is insufficient when they come to the next questions.
  • When the idea of political correctness is introduced, respondents are led to see it as something that limits the police’s effectiveness. Respondents are forced to choose between: “The police feel like they’re on my side with my priorities and interests at heart” and “The police increasingly feel as if they have their own politically correct agenda which does not match my interests”. This assumes political correctness is opposed to respondents’ interests: the choice is between being politically correct and acting in respondents’ interests. While it’s normal for these questions (known as polarities) to force respondents to choose between extreme positions, this question conflates two different debates (are the police politically correct? is that good or bad?) and in doing so leads respondents to think negatively about political correctness in policing.
  • In a series of statements, which respondents can agree or disagree with, six of the seven are phrased where agreement gives the answer that We, The People presumably wanted. This is bad polling practice. If you really want to measure public opinion you ask a question that presents both sides of an argument equally, then allow respondents to choose which they are closer to. Or if you really have to ask agree/disagree questions, the questions should be balanced overall so you’re not pushing a particular argument and you can compare skewed questions against each other.

In fairness, the poll is less bad than the previous one: the wording of each individual question is less skewed this time. But the order and overall balance of the questions still add up to a loaded poll.

A good guide of a fair poll is that you shouldn’t be able to guess the view of the organisation commissioning the poll from the questions. This clearly fails that.

It is also arguable the poll fails the Market Research Society’s Code of Conduct, which says (33d) “Members must take reasonable steps to ensure … that participants are not led towards a particular point of view”.

Amusingly, despite the leading questions, the poll still got some results that the Mail choose to ignore, including:

  • 48% agreed “The police need to act with political correctness as it encourages acceptance and decency in society”, with 32% disagreeing (strangely, this was missed off We, The People’s press release).
  • 59% agreed “Hate crime is a blight on our society and the police are right to try to tackle it” compared with 24% saying “The concept of hate crime is well intentioned but threatens Britain’s heritage of free speech and open expression”.

 

Emergency podcast: Bye bye Boris & what happens next? Polling Matters

Posted in Politics, Polling Matters on July 9th, 2018 by Leo – Be the first to comment

On a special episode of the Polling Matters podcast, Keiran and I discuss a momentous 24 hours in Westminster that has seen both David Davis and Boris Johnson resign. We ask what happens next and look at what polling of Tory members by YouGov tells us about the future direction of the Conservative Party and who might come to lead it.