Opinion polls

by Chris Bertram on July 26, 2005

Since the bomb attacks in London there have been a number of polls which, among other things, ask British Muslims whether or not they thought the attacks were justified. This then provides material for op-ed columnists and bloggers to scale up the number so as to argue that there are {insert large number} Muslims who are prepared to back the terrorists. Having looked at the detail of the latest poll, from “ICM as reported in the Guardian”:http://image.guardian.co.uk/sys-files/Politics/documents/2005/07/26/Muslim-Poll.pdf (pdf), I’m sceptical about any such conclusions given the strange combination of views apparently endorsed by respondents in Table 8.

Out of 500 Muslim respondents, 26 said the bombings were justified. Of those 26 bomb-justifiers, 7 declared they would vote Conservative if there were a general election tomorrow, 12 were potential Labour voters and just 2 backed the Liberal Democrats. Go figure.

{ 15 comments }

1

Otto 07.26.05 at 6:34 am

Ah the Scheissliberalen! All talk and no action.

2

David Deans 07.26.05 at 7:59 am

Funny, I was just posting on exactly this topic.

3

Kevin Donoghue 07.26.05 at 8:01 am

Out of 500 Muslim respondents, 26 said the bombings were justified.

Actually the question was whether any further attacks are justified. Different question, scarier answer.

What caught my eye was the fact that 14 of the 26 were female, which I didn’t expect.

4

Bamacrat 07.26.05 at 8:27 am

I agree. The number of women that thought the bombings were justified is surprising.

5

Tim 07.26.05 at 8:37 am

One shouldn’t make too much of results with a base of 26 respondents: the margin of error is about 19.2%, 19 times out of 20 (this presumes a random sample, of course).

This leads me to wonder what exactly one uses as a sampling frame for British Muslims, anyway — all the names in the telephone directory that “sound like a Muslim name” by some pollster’s definition? There’s potentially a lot of non-randomness in this poll that casts doubt on the findings irrespective of the actual results.

6

David Deans 07.26.05 at 9:12 am

Census data? I guess you can guess by where people live, as many Pakistani communities are still very nebulus and live in concentrated areas, such as Small Heath or Dudley. Just ring around until you get your willing Muslim…

“So Mr Akram, since the London bombings do you feel more or less likely to blow up the Houses of Parliament? Answer on a sliding scale, with 5 if dead keen, 3 if not sure, or 1 if you’d haven’t got the time this week and you’re busy with your children…”

7

Tim 07.26.05 at 9:33 am

I guess you can guess by where people live, as many Pakistani communities are still very nebul[o]us and live in concentrated areas, such as Small Heath or Dudley.

That’s really a dressed-up sample of convenience, nothing more. As a sampling strategy that would give a Muslim living in, say, Aberystwyth a zero probability of being polled.

Guessing by where people live means introducing a lot of non-randomness, and consequently dodgy poll results.

8

Cryptic Ned 07.26.05 at 9:49 am

There’s a Muslim living in Aberystwyth?

9

Ben 07.26.05 at 10:00 am

10

aber 07.26.05 at 10:01 am

There’s a Muslim living in Aberystwyth?

Well, Aberystwyth is a typical university town with (a) a left-wing student body and (b) a considerable number of international students, especially in its large Department of International Politics, some of them Muslim.

Also, Aberystwyth is well known for its radical activism — just ask the Queen!

But then again, if Aberystwyth produces any terrorists, they are more likely to be Welsh nationalists…

11

David Deans 07.26.05 at 10:09 am

Tim – yup, agree. Inless you only wanted to know about people who lived in Small Heath, that would be useless.

It appears ICM took their sample from previous respondees where the individual has said they were Muslim, see here.

12

Tim 07.26.05 at 10:44 am

David — thanks for the link to the ICM methodology.

Re-interviewing people from a previous survey isn’t that objectionable, IMO, though the 45 “referrals” to boost the sample size to 500 puts a dent in ICM’s methodology. I would have stuck with the original 455 respondents.

“Thank you for participating in our survey. Do you know any other Muslims who might be willing to answer our survey, the results of which may fuel right-wing Islamophobia?”

13

soru 07.26.05 at 2:43 pm

Some people think that information can dispell ignorance, and that that can be a good thing.

Some people, evidently, don’t.

soru

14

Robin Green 07.27.05 at 7:09 am

I want to try to put this apparent support for terrorism from a few British Muslims in some sort of perspective.

How many American Republicans (or, hell, even Democrats) have come around to the idea that torture of suspected terrorists, or terrorist sympathisers with information, is acceptable in some circumstances?

I would argue that supporting torture of potentially innocent people is actually morally equivalent to supporting bombing innocent people. Where is the outrage about this?

It seems that it is not OK to have a rational debate about the morality of Islamist terrorism, but it is OK to have a rational debate about the morality of torture. I don’t really understand why there is this difference.

It is, of course, human nature to react more sympathetically to member of your perceived “tribe” (in this context I mean non-terrorist white Westerners [if torture is not classified as terrorism, which is in itself debatable]) than to members of another perceived “tribe” (in this context, meaning e.g. Muslims). This to some extent explains it. But still. It does not make it logical.

15

Robin Green 07.27.05 at 7:24 am

I must correct myself:

or terrorist sympathisers with information

That should of course read “people suspected of being terrorist sympathisers with information”. Countries like the US and Saudi Arabia do not, of course, convict people in a fair trial of being terrorist sympathisers before taking them away to be tortured. That’s just not how it works.

(Saudi Arabia does not have fair trials, period, and the US has often not bothered with trials at all for people they suspect of having information relevant to terrorism or insurgency.) So it is in fact quite likely that some people who were thought to be terrorist sympathisers by these outlaw regimes, were in fact completely innocent.

Comments on this entry are closed.