Comments on: Tory lead faltering in two new polls

By: Anthony Wells

Anthony Wells — Mon, 28 Jan 2008 21:31:36 +0000

Could be. Stranger things have happened.

By: Mark Senior

Mark Senior — Mon, 28 Jan 2008 17:33:54 +0000

Anthony , no I don’t think that is the explanation . in every age group the one who lead in the previous poll now trails by a similar amount , I would guess their figures have been transposed .

By: Anthony Wells

Anthony Wells — Mon, 28 Jan 2008 17:06:35 +0000

Frederic – ICM polls are sampled and weighted to to correct proportions of people from each region. 9% of their sample is made up of people from Scotland, the same as every other pollster.

Mark – I suspect the something wrong is just very low sample size. Take out the 100 people who were won’t votes or Don’t Knows and there must be all of 240 voting intentions. In each age bracket the numbers are embarrassingly low.

By: Mark Senior

Mark Senior — Mon, 28 Jan 2008 16:53:05 +0000

The detailed data for the ICM and 2 Yougov polls are now on their websites .
The Yougov Mayoral poll sample size is rather low only 339 . There is something wrong with the detailed data compared to the last poll . In the previous poll Boris led Ken in the over 55′s but this time Ken leads Boris by a substantial amount . The other age groups also have an apparent clear volte face in who is leading .
The ICM comparison with how people voted in 2005 is most interesting . LibDem retention is the highest for some considerable time . Just 8 switchers to Labour with 7 switching in return and just 9 to Conservative with 4 switching from Con to LibDem . Cleggy seems to have won back the majority of those who had recently been saying they would vote Conservative .

By: Frederic Stansfield

Frederic Stansfield — Mon, 28 Jan 2008 16:49:41 +0000

The YouGov poll seems to give the Others 10%, rougly in line with most recent polls. By contrast, the ICM poll appears to squeeze “Others” to 8%.

As a large proportion of Other respondents is made up of SNP supporters North of the border, I suggest that anybody wanting to see why the ICM poll might be out of line should enquire about the geographical distribution of the people sampled and the breakdown of responses by “region”.

The differences between the YouGov and ICM polls appear to be within sampling error, however.

This post reflects previous points I have made on this site about issues raised by the Other category for interpreting national opinion polls.

Turning to the causes of the reported figures, there currently appears to be little movement in the polls, and these two appear not to be out of line. The raise in Labour figures might be explained becuase the Westminster executive is now in operation again after the New Year, even if not doing anything very inspiring. On the other hand, the opposition are probably holding any major initiatives for a better moment.

The Liberal Democrats might have hoped for a better “bounce” following the election of their new leader. In particular, Clegg’s first weeks at Prime Minster’s Question Time seem to have had little impact (certainly not by comparison with Vince Cable). As first impressions are important, these polls may not be good news for him.