Skip to main content icon/video/no-internet

Opinion polls are sample-based and interview-led quantitative surveys designed for collecting information on the attitudes and behaviors of a given population on a certain set of social, political, ethical, or economic issues. In both historical and methodological terms, opinion polls are closely linked to market research, of which they represent a subsequent extension to the study of public opinion. It was consumer research that first used sampling and interviewing to analyze the habits, interests, needs, and preferences of consumers. Since the beginning of the twentieth century, companies have employed quantitative surveys, alongside qualitative research techniques, for such diverse purposes as probing consumer perceptions or revealing consumer trends, gathering consumer feedback or monitoring consumer moods. A broad majority of polling organizations continue to base their business mainly on market and consumer research, while public opinion polls, preelection polling, or exit polls represent only a marginal share of their activities. More specifically, the most direct historical antecedents of opinion polls are to be found in the market research, which some United States agencies began to carry out on behalf of businesses or publications from 1910 onward. The purpose of this research, which was already based on the principles of sampling and interview methods, was to provide their clients with information about the characteristics and preferences of their customers or readers. The major pioneers of modern opinion polls, George Gallup and Elmo Roper, came from the field of market research in the United States. Both published the results of their respective opinion polls for the first time in 1935. The magazine Fortune published Roper's poll, relating both to the consumption of cigarettes and the possession of motor cars by their interviewees and to their opinions on the subjects of taxation or the role of government in creating jobs.

Methodological Issues

The notion of representative sampling is one of the methodological pivots of opinion polls. A sample is a group of individuals who are selected as representative of a given population (citizens, voters, consumers, youngsters, housewives, etc.) through random extraction (probability sampling). According to the mathematical theory of probability, assigning each member of a population with an equal probability of being selected is a sufficient condition for obtaining a final sample that is relatively similar to the whole population, with a predictable degree of approximation, which depends on the sample size. Due to an inherent statistical bias known as sampling error, the expected gap between the values in the sample and in the population (or the confidence interval) is, for example, around 3 percentage points for a sample of one thousand interviewees—the standard size in contemporary opinion polling. This means that an opinion expressed by, say, 28 percent of interviewees more realistically reflects the opinion of a proportion of between 25 and 31 percent (28 percent plus/minus 3 percentage points) of the observed population. In addition to probability sampling, another technique, known as quota sampling, is used, particularly by non–Anglo-American commercial pollsters, to ensure certain sociodemographic variables (sex, age, region, occupation, etc.) with exactly the same proportion in the sample as in the general population. Although it may be effective for practical purposes, the quota sampling technique (or nonprobability sampling) is regarded as arbitrary from a statistical theory perspective, since it is incongruent with the law of probability. However, since the strict theoretical assumptions required for probability sampling are systematically violated in real-world practices (i.e., low response rates among the originally sampled individuals), a mixed method combining probability and quota sampling can be considered as standard procedure in modern public opinion polling.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading