In a way it’s about cause and effect.

If two things happen at the same time it may be that they are linked but it might be a coincidence.

If however you see a change and there is an associated event that could be an explanation then it is more likely ( though not certain) that there is a link.

The danger is when two things happen at the same time and we assume that there must be a link.

Peter.

Peter makes a very valid point about the regional breaks being useful for identifying how different policies / tactics play in different parts of teh country. Equally, his point about volatility is clear. One simply has to accept that there is a higher margin of error in the smaller samples.

On the other hand, if there are slow but steady trends over time, then this information is useful and can be deployed in an election campaign.

As Peter noted, a 50% shift from one month to another needs to be taken with a inch of salt – unless the figure then stabilises at the new leevl – in which case one should look to see if there is any likely causal event which could shape the changed response.

If one were to see a “jump” in SNP & Tory results, with corresponding falls in Labour and LD scores, in Scotland over the next few polls, then, even if the numbers settle back after a while, it would be possible to deduce this as an effect of the Budget debate. If on the other hand the reverse were true, then I think we would be scratching our heads for an explanation.