As I said in my post last night, one explanation for the differing results given in the post-debate opinion polls was that they could be asking different questions. Having had a chance to look at some of the reporting of the polls, it seems they were – or, at least, they asked a variety of questions but the headline results given were in response to different questions.
So, we have two polls asking about who performed best and two asking who people thought ‘won’. While these may seem similar, they’re actually asking people to judge different things. ‘Performing best’ is asking for a judgement on observable factors, while given that there was no objective scoring available, ‘winning’ is perhaps expecting people to think about how other people would have judged the debate.
So, was there a difference in the results? Here’s the results from the pollsters who asked the ‘winning’ question:
Both of them put Cameron and Miliband ahead, and effectively tied, Farage just behind them, Sturgeon in the mid-teens and Clegg, Wood and Bennett in single figures.
Now, those who asked about best performance got these results:
So, we may have a difference in the results based on the question asked. Cameron and Miliband both do worse (but perform closely together) in performance questions compared to winning ones, while Sturgeon does better when people are asked to judge performance. She gets the best result of the night for anyone across all the polls in YouGov when people are asked to leave aside their own political preference and just judge performance.
This is based on a small sample of four polls and just one debate, so there’s no way the results are conclusive, but there does appear to be an interesting difference between the results from the two questions. When asked about winning, it’s possible some people are more likely to think of it being a battle between Cameron and Miliband than they are when asked to consider who gave the best performance of the night. Election coverage is shot through with the idea that only one of those two can ‘win’ the election, so it’s not unsurprising that a question about winning gives them more support than a question about performance does.
However, with neither of them having a decisive lead over the other and them not being too far ahead of the others, it suggests there isn’t much expectation of a decisive victory for either of them in the election. Even when asked to make a forced choice between the two of them, the results were evenly split, and going by that evidence, we’re set for a very inconclusive election result.