But poll junkies will have to wait for YouGov

I was expecting to be leading this morning on the November YouGov poll which invariably features in the Daily Telegraph on the final Friday of each month. Well it does not appear to be there but for those those who need their regular “fix” of polling information we do have the full dataset from this week’s ICM survey for the Guardian.

The main point from these numbers is that they reinforce the trend from recent polls showing that the move to the Tories has been biggest amongst men than women.

With males the split was CON 36%: LAB 34%: LD 20% while with the females in the sample the shares were CON 38%: LAB 31%: LD 22%.

But ICM did not find the same dramatic difference as Populus earlier in the month when it asked the named leader question – how would you vote if it was Cameron’s Tories, Brown’s Labour and Campbell’s Lib Dems.

On a cautionary note what the full data also shows is that when you look at the standard voting intention and contrast it with the now almost inevitable named leader question you are not comparing like with like. It is very easy, as we saw in the Guardian on Wednesday, to seek to compare the two findings but they are calculated on a different basis.

The named leader numbers are based on the full sample while the main voting intention figures are adjusted in accordance with how likely it is that respondents say they will vote. Given that this measure usually helps the Tory number a turnout linked adjustment would probably show a bigger gap – maybe 1-2% more.

On the other hand the pollster does not use the “spiral of silence” adjustment with the named leader question and this in recent months has sometimes been giving Labour a slight boost.

This is a bit technical but it is very easy to get drawn into comparisons which are not strictly accurate and I acknowledge that I am as guilty as anybody.

Mori note. The Sun survey that appeared on Tuesday was not from Mori’s main poll for the month. That should be out this weekend.

Mike Smithson