One-Party State (see Single-Party System)
Opinion Poll
Opinion polls are measures of public attitude, on any sort of issue, carried out
by professional polling organizations whose main business is usually in market
research. In themass mediaa typical opinion poll is a measure of thevoting
intention of the electorate, of how the voters rank politicians and parties, and
of what preferences the electorate has among various policy alternatives; these
polls are taken with increasing frequency and assume greater importance as an
election approaches. The techniques of such polling are complex; essentially
there are two aspects, sampling and questionnaire design. The most accurate
and reliable sampling designs are seldom used for the sort of political opinion
polls published in the media both because they are expensive, and because they
cannot produce results as quickly as is often required. A typical opinion poll
will have a sample of between 800 and 1,200 voters selected by a system known
as ‘quota sampling’. With this method interviewers are sent to perhaps 50
locations, usually parliamentary constituencies, which are chosen randomly.
They are each required to interview a sample constructed according to gender,
age and class to provide a cross-section of the population. Although accuracy
increases with larger samples, it does not do so proportionately, so that
doubling the size of a sample does not halve the margin of error. Any answer
in a sample of 1,000 is liable to a margin of error of plus or minus 3%.
Questionnaire design is both more complex, and less calculable in its effects.
It is known that the wording of a question can affect the probability of a
particular answer being given considerably, but so can something as unpre-
dictable as the ordering of questions. Great efforts are made to make the
questionnaires as neutral as possible, but it is hard to be sure this has been
achieved. During the last decades of the 20th century it began to be clear that
election-time polls in the UK systematically under-measured Conservative
strength in ways that could not be accounted for by sample design; instead
there appeared to be a need further to refine question wording. It needs to be
remembered that, although during an election the polls will be entirely about
politics, regular monthly polls are often added to lengthy consumer attitude
questionnaires designed for the advertising industry, with possible effects on
responses.
As well as the question of accuracy, concern is often expressed about the
possibility of the publication of polls actually changing voters’ attitudes. There
is some evidence that the publication of poll results does affect attitudes,
through a variety of psychological factors. For example some people have a
tendency to want to be on the winning side, and will shift their preference to a
party, or policy position, which seems to be gaining support. Meanwhile,
One-Party State