h1

Guess which of these polls had mobile users in their samples?

November 3rd, 2008


    If Obama wins big will UK pollsters have to change?

The chart is from Nate Silver’s excellent fivethirtyeight.com and shows the Obama leads from the major national White House polls colour-coded according to whether the firms do or do not include cellphone users in their samples. You guessed it – the ones in yellow did while the one in grey restrict their samples to people who could by standard land-lines.

There has long been a debate on both sides of the Atlantic over the range of telephone numbers that are used in political polling and whether the increasing proportion of people who only use their mobiles could be distorting samples. In the UK the normal approach is to take phone numbers that are listed and then for the computerised dialler to randomise the final digit.

My guess is that if the final result is closer to the top of the chart – a big Obama victory on votes – then we’ll hear a lot more of this effect in the run-up to the UK general election.

One factor in the US is that large numbers of mobile phone users have to pay both to send calls and to receive them. So if you receive of polling call on you mobile then it will cost you. But many price plans bundle in free weekend minutes and Nate is speculating over whether final polls carried out over Saturday and Sunday might reflect a different pattern with mobile users being more willing to take part.

    Whatever the polling picture supports the widespread view that Barack Obama is heading for a big victory tomorrow – maybe even a landslide.

I’ll be doing a round-up of the latest electoral college voting spread betting when the markets open during the day-time.

For a full round-up of conventional White House Race betting click the panel below. This will take you to PB’s micro-site which is run by our co-sponsor, Bestbetting – the UK’s leading provider of live betting from a range of bookmakers and exchanges.

Mike Smithson






Comments are closed.