Ipsos polled the Sir Humphrey questions
Ipsos report:
In a famous Yes, Prime Minister episode Sir Humphrey Appleby once explained to Bernard Woolley how you could get contradictory polling results on the same topic – in this case the reintroduction of national service – by asking a series of leading questions beforehand and asking the key question you want to know about in a certain way.
But what would happen if we asked Sir Humphrey’s questions today? To find out we asked 1,000 British adults the first set of questions that were positive about national service and 1,000 British adults the second set that were negative. Below is a comparison of the results. You can see that it is indeed true that you get different results on the level of support for the reintroduction of national service based on the way you ask the question and the questions you ask before it.
The results are:
Ipsos point out:
First of all, to be clear, we would never ask questions on such a topic in this way. These are taken from a comedy sketch, a sketch that works because of the obvious absurdity of what Sir Humphrey is suggesting. At Ipsos, we take great care not to ask leading questions. Questionnaire design is a key part in how our researchers are trained. Our professional reputation is based on providing quality data and acting with integrity. It is why clients come to us.
Related to this, it is worth pointing out that Sir Humphrey’s central allegation – that polling companies don’t publish all of the questions they include in surveys – is not allowed under British Polling Council rules. Under these rules, polling companies have to be transparent. They must publish all relevant questions asked in a poll in the order they were asked. Interested parties can then reasonably disagree on question wording and so on in an open way. In many ways this acts as something of a public peer review of survey results.
The same in true in NZ. All relevant questions must be published, and for political polls the primary voting question should be asked before any other question which could influence it.