Friday, September 19, 2008
Odd Attacks and Interesting Data
A poll published by the Indepedent which shows Gordon Brown as almost the most unpopular cabinet member and that 54% of Labour activists would prefer someone else to be leader has received a lot of publicity as well as scathing attacks on the Ministry of Truth blog and Iain Dale's Diary. MoT's attack on LabourHome notwithstanding, the basis for dispute is "the poll had been commissioned by the Indy... where’s the demographic information that pollsters collect as standard in order to balance their polls and ensure their statistical validity and what’s the margin of error on the numbers......" Well it is not unusual for media organisations to commission polls and not state the fact, even when it is the phone in poll, and the article does say "an exclusive poll for The Independent". But the main problem is validity, but it was an online poll of an audience that is largely unknown. The invitation was to "With the Westminster Villiage obsessed with the question of leadership of the Labour Party, we thought it's time to find out what the grassroots thinks. Click here to take part in the Labour Grassroots Survey and we'll publish the results in the first days of Labour Party Conference." So the respondents, from a total population of unknown numbers, could be non-average (unrepresentative) just because the link was only there for a matter of days. There were 788 members, all of whom must have been checked to see if they were supporters as "Non Labour supporters who responded to the survey were stripped from the results" though this may have still been skewed by opponents claiming to be supporters. However, a serious question is how you get to the engaged and active supporters of a party. Chances are, in the digital age, they are online and accessing party communication and taking part in social networks. But is it really so wrong to do a poll online? It is impossible to offer the normal caveats or statistics beyond the number of respondents, and perhaps The Independent should have stated how it was conducted, and yes sponsorship should have been mentioned by Labourhome in the link as perhaps some supporters were a little too honest (though it does say data will be published), but are all the attacks really justified and why are the opposition attacking when really it is better for them and for Labour. As I understand it, Labourhome is independent of the party leadership and perhaps the one place where inconvenient truths can be aired, such things should be read by the party leaders as it may jsut enhance the connection between the party and their foot soldiers. For PR purposes, and if I was advising Brown, I would face this head on in the speech this week!