Can we trust today's Times poll?

Big doubts are now attached to this morning's Times survey which gave David Davis a surprising lead amongst Tory voters.  The result was based on a sample of only 122 voters.  Populus started with a sample of 1,500 for their headline voting intention number but then split the sample into two and asked 750 people different questions.  Such a split is usually done to save the client (in this case The Times) some money.

The 750 or so voters of which the Tory leadership question was asked included only 122 Tory voters (once the ‘won’t votes’ and ‘don’t knows’ etc are excluded).  Therefore: The figures splashed by The Times have a massive sampling error.  Populus have been enirely open with the limited size of the sample on their website but has The Times been irresponsible with its coverage?

Comments

[michael]

No. Because away from that dreadful headline - the figures are actually quite good for Cameron.

As, I said - Davis is now more unpopular with voters generally than when they didn't know who he was! The reverse is true of Cameron.

[petersmith]

"The figures splashed by The Times have a massive sampling error." I think you mean 'margin of error'. It may be a perfect sample.

On these figures, Cameron could theoretically be ahead - so yes, it was wrong of The Times to splash on it. Wrong and absurd.

What 'margin of error' means is that there is a 95% likelihood that Davis is between 37% and 63%, and Cameron is between 24% and 50%.

HOWEVER, it is not equally likely that the true figures are anywhere within that range; they are MORE likely to be close to those given.

So, to be completely fair, one would have to say that the poll figures are still interesting and probably good for DD - just not good enough for a front-page splash.

But then again, do you expect serious reporting from The Times? As you well know, there's a 50% chance that anything The Times says is rubbish.

[Rupert Murdoch]

"Davis races ahead in Tory poll" - it is unfair because it is misleading! Conservative voters are not Conservative members (although I hope both groups are voting Conservative!)

The Times has been deliberately naughty.

[Mark Fulford]

The Times lost it's reputation for thoroughness long ago. This headline wasn't irresponsible because it's hard to see that it has done anyone much damage. It was simply headlines and sales on the cheap.

[James Hellyer]

Even with the sampling margin of error (8%) that still leaves Davis and Cameron closer than any other poll in the run off.

But it isn't a poll of Party members that's what is so annoying James.

[malcolm]

Not on subject but Blair has lost the 1st of the 90 day detention votes.Congratulations to all who voted against this proposal.The fact that the Sun will be hacked off is an added bonus.

[William Randolph Hearst]

"The Times lost it's reputation for thoroughness long ago."

Exactly. That must mean their endorsement of Cameron can be junked too.

[Mark Fulford]

Throroughness, not opinion.

[Jack Stone]

Thank god for that. Thought Conservative supporters had taken complete leave of there senses for one moment

[Wat Tyler]

We've actually had much smaller samples in previous leadership polls. What about the one that showed DD behind DC in the North- on a sample of 50?

But quite agree about the Times. It's turning into a poor imitation of the Grauniad. I must cancel it...the trouble is, I'm gradually being forced to cancel everything.

[James Hellyer]

Well Wat, I suspect Camp Cameron would complain that "that was different!"

[malcolm]

I know how you feel Wat,but remember we have to live with the media we've got just as Labour had to in the 1980's.Blair great skill was to get some of the media onside who were previously against it.Sadly we have to the same.As soon as the media start to believe that we are serious contenders then I predict some sections of it will come around to our way of thinking.

[Andrew]

I like the Times. Know thine enemy, and all that...

[loyal_tory]

Please correct me if I'm wrong but while the margin of error for the Populus poll is 9%--meaning Davis could have anything from 59% to 41%, and Cameron anything from 26% to 44% (Davis could be ahead by up to 33 points and Cameron by up to 3 points), the latest YouGov poll had a margin of error of 8% (excluding don't knows)--which means that Davis could have anything from 40% to 24% and Cameron between 60% and 76% (anything from a 20 point to a 50 point lead).

It is important to remember that both latest sets of poll numbers, the one of members (YouGov/Telegraph) and the one of Tory voters (Populus/Times) had very small sample sizes and consequently a high margin of error.

Also Andrew Cooper of Populus has a post on politicalbetting.com arguing why Tory voters could be considered an unimportant sample (stastically few will be members and it doesn't matter what they think since they vote Tory anyway). That may be true but it could be argued that members, as opposed to activists, are likely to be closer in their opinions to Tory voters. And the YouGov sample may over-represent activists as a proportion of the total members. I wouldn't be at all surprised if activists were more pro-Cameron and non-active members were more pro-Davis but with the polls we have to work with it is incredibly difficult to tell!

[Cllr Iain Lindley]

The YouGov survey had a much larger sample IIRC.

Surely the margins for error only apply if you have a representative sample in the first place. I doubt 120 people is a representative sample of anything.

[Guido Fawkes]

This morning's Times survey headline was based on just 61 votes! Head of Populus, Andrew Cooper, told the punters over at PoliticalBetting.Com that he "would not have put the headline they (the Times) did above the findings ... Cameron has a (much bigger) lead when it comes to which of the two is judged more likely to win an election, unite the party, get it in touch with ordinary people etc. And all the poll questions on the Tory leadership indicate that Cameron attracts more support than Davis does from those not currently Conservative supporters."

Bookies raced to re-price odds when the full details of the survey were posted on the Populus website, with Davis odds collapsing back out from 3/1 to 5/1, Cameron's odds shortened back to hot favourite. Cooper is blaming The Times for the headline. Lets be honest, a second abberant poll for Populus this year would be embarrassing, given they predicted a landslide for Labour at the general election.

[Andrew]

The Times is as much of a tabloid as the Daily Morons (aka Mail/Mirror) - I'm only waiting for it to complete the process and formally merge with the Sun.

[loyal_tory]

Sorry my mistake earlier, it was the ICM poll (that put Cameron on 76% and Davis on 24%) that came out before the YouGov poll that had a margin of error of c.8% (sample: 215 people). The later, YouGov poll (Cameron 68%, Davis 32%) had a margin of error of c.4% (sample: 621 people). So YouGov in fact has a lower margin of error than the numbers we've seen quoted from ICM or Populus via The Times today.

[carol42]

If Cameron does not face Paxman then that is it for me. If he can't cope with that type of interview then he will be destroyed by Blair and Brown easily.

[Wat Tyler]

Hope Populus have got some other clients, cos I don't expect they'll get much more from Murdo's guys...don't they understand the relationship between customer and supplier?

[john Skinner]

The Times has a vested interest in damaging the stronger of the two Tory leadership contestants. Who can trust a newspaper that employs Alastair Campbell and Tom Baldwin?
Incidentally, Cameron is right to keep Paxo waiting. Why should the media call the tune?
Paxman is a lazy, self-indulgent bully. He appears to do very little preparation, but relies upon eye-popping intimidation and browbeating. Nobody comes out well after a shouting match with Mr Nasty of Newsnight.
Cameron should agree to an in-depth interview with John Humphrys, a much better- prepared and serious interviewer.

[James Hellyer]

"Why should the media call the tune?"

Yes, let's wait until he's elected to see if he's up to the job. What a good idea!

[Peter C Glover]

I have been giving some thought of late to the theory of relativity and Tory leadership candidate David Cameron.

The question is this: is the vision-less, policy-lite Cameron, political speaking, only relatively clueless or completely clueless?

Or is this just another of life's eternal mysteries?

[wasp]

I don't think any of the polls have been trustworthy, so I'm going on speaking to a lot of members who seem to be Cameron by a whisker.