Isn’t it all about how the questions are framed?

The above clip always makes me chuckle and although it’s from a comedy show and is perhaps a bit over the top it’s a useful reminder about the hazards of non-standard polling questions.

So what are we too make of the poll findings from the ComRes survey that the Indy is leading on this morning.

The first non-voting related question asked whether respondees agreed that “A Conservative Government would mainly represent the interests of the well-off rather than ordinary people” to which there was 52% to 44% agreement

This was followed directly afterwards with “A Labour Government would protect frontline public services such as health and education better than a Conservative Government would” which produced a split of 47% to 46% in agreement.

Finally, ComRes asked without referring to any party whether the sample agreed that “The threshold for paying inheritance tax should be raised to £1 million”. To this 55% said they agreed and 38% said they didn’t.

Surely, as Richard Nabavi suggested on the previous thread a more neutral approach would have been to ask:

“Which party has the better policies for health and education?”

“Which of the Labour and Conservative parties is more appealing?”

At this stage the party strategists want data that will help them frame their approaches better and I don’t think the ComRes/Indy formulation is helpful. Taking too much notice of any of the responses, including the IHT one, could lead to flawed conclusions.

Mike Smithson