MEASURING PUBLIC OPINION| 135population, or group of people, such as the residents of a specific congressio-
nal district, evangelicals, senior citizens, or even the nation’s entire adult popu-
lation. For la rge g roups such a s t hese, it wou ld be i mpossible to su r vey ever yone.
So surveys typically involve samples of between a few hundred and several
thousand individuals.
Mass Surveys
One of the principal attractions of mass surveys is that they can in theory pro-
vide very accurate estimates of public opinion for a large population by using
relatively small samples (see Nuts and Bolts 5.1). For example, while polls taken
very early in a presidential campaign (such as a year in advance of the general
election) are poor predictors of the ultimate outcome, polls taken after the cam-
paign has gotten started, when both party’s nominees are known, provide very
good predictions of who will win the election and how many votes that candidate
will receive.^42
Large-scale surveys such as the American National Election Study (NES),
conducted every election year, use various types of questions to measure citizens’
opinions. In presidential election years, participants in the NES are fi rst asked
whether they voted for president. If they say yes, they are asked which candidate
they voted for: a major party candidate (Barack Obama or Mitt Romney in 2012),
an independent candidate, or some other candidate.
Another kind of survey question uses an issue scale. For example, respondents
might be asked about the government’s role in protecting morality or their own
liberal–conservative ideology. Two opposing statements are given for each topic,
and respondents are asked to agree with the one that comes closest to their views,
including options in the middle of the two extremes.
A typical survey includes a hundred or so questions asking respondents to
evaluate issues and candidates, along with questions about the respondents’
age, education, and marital status. Some surveys conducted by candidates or
political parties are shorter, focusing on voter evaluations of specific candi-
dates and the reasons for these evaluations. In the main, the length of a survey
reflects the fact that interviewing people is expensive, so there is a trade-off
between learning more about each respondent’s opinions and maximizing the
number of respondents. The fewer questions asked, the more people can be
interviewed.
PROBLEMS IN MEASURING PUBLIC OPINION
While measuring public opinion seems an easy task—just fi nd some people and ask
them questions—it is actually very complicated. The problems begin with gather-
ing an appropriate sample and are compounded with issues such as question word-
ing and the nature of public opinion itself. As a result, survey results must be read
carefully, taking into account who the respondents are, what opinions they are
being asked to give, when they are being surveyed, and what mechanism is used to
ask the survey questions.
mass survey A way to measure
public opinion by interviewing a
large sample of the population.
population The group of people
that a researcher or pollster wants
to study, such as evangelicals, sen-
ior citizens, or Americans.
sample Within a population, the
group of people surveyed in order
to gauge the whole population’s
opinion. Researchers use samples
because it would be impossible to
interview the entire population.