It's That Poll Again The principal polls for sampling public
opinion are the National Opinion Poll, published in the Daily Mail, and the Gallup Poll, published by the Daily Telegraph. With no more by-elections to come in this Parliament, it is all the more deplorable that they are once more wildly out of step. I share Harold Wilson's opinion that the NOP is the more reliable and transferred my allegiance long ago. So you must allow for a certain prejudice as I examine the efforts of Dr. Durant of the Gallup Poll to prove that quota sampling is superior to the random sampling used by the NOP. In this, by the way, he goes flatly against his chief, for Dr. Gallup, in an interview in the Home Service on May 7, said, 'Well, there are many systems, and I think the best system is one that is described as "random sampling" . . .' and goes on to give convincing arguments against the quota method. The two Doctors can sort that one out for themselves.
The absurdity of the present difference (NOP gives the Tories an excellent chance, and Gallup has been forecasting a majority of about one hundred for the Socialists) has led Professor Eysenck (yes, same chap) to suggest that these exercises in soothsaying should be examined by 'an independent academic organisation' set up and financed by the Social Science Research Council. Heaven forbid.
The new line of defence now put forward by Dr. Durant is that quota sampling is closer to the Registrar-General's figures than is random sampling. As quota sampling is, in fact, sup- posed to be based on these figures this is not very surprising. But the comparison seems irrelevant. What one wants for political forecasting is a sample of the electorate, not of the population, and particularly in the younger age groups this is by no means the same thing: many of those between twenty-one and twenty-five do not, in fact, come on to the register as early as they could.
In short, I find the new apologia just as uncon- vincing as I did the last and I return to the criticism I made of the Gallup Poll in this 'Notebook' on March 20: 'There is a fever in this chart, but it is hard to find it mirrored in the country.'