Demographic Measure

Demographic measures are questions that allow pollsters and other survey researchers to identify nonopin-ion characteristics of a respondent, such as age, race, and educational attainment. Demographic measures typically are used to identify key respondent characteristics that might influence opinion and/or are correlated with behaviors and experiences. These questions are usually found at the end of a questionnaire. Reasons for this are (a) to engage or otherwise build rapport with the respondent by asking substantive questions of interest earlier in the questionnaire; (b) to lessen the likelihood that asking these personal questions will lead to a refusal to continue completing the questionnaire (i.e. a breakoff); (c) to prevent priming the respondent; and (d) to allow the respondent to answer the core questions before possibly boring him or her with the mundane demographic details.

Demographic measures are important because numerous studies have demonstrated that opinions are formed primarily through an individual's environment. This environment socializes us to think and behave in accordance with community norms and standards. As a result, by identifying these demographic measures, pollsters are better suited to understand the nature of public opinion and possibly how it might be formed and modified.

Demographic measures are also very important because they allow researchers to know how closely the sample resembles the target population. In a national sample of U.S. citizens, for example, researchers know what the population looks like, demographically, because the federal government conducts a census every 10 years and updates those data annually thereafter until the next census. As such, researchers know the percentages of the population based on race, gender, age, education, and a whole host of other demographic characteristics. A simple random sample of the population ideally should resemble the population, and demographic measures allow researchers to see how well it does. For example, because survey nonre-sponse often correlates with educational attainment, most surveys of the public gather data from proportionally far too many respondents who earned college degrees and far too few respondents who did not graduate from high school. Knowing the demographic characteristics of the sample respondents (in this case, [Page 186]educational attainment) allows the researchers to adjust (weight) their sample to the known population characteristics. This can be done with greater confidence and accuracy if the wording of the demographic question in the survey matches the wording of the question for the same characteristics that was used to produce the universe estimates (e.g. the wording used by the U.S. Census).

The length of the questionnaire often limits the number of demographic questions asked. Accordingly, demographic measures must be carefully selected to best allow further analysis. There are a number of standard demographic questions that are nearly always asked, including questions about age, gender, income, race, Hispanic ethnicity, and education. Questions designed to identify these characteristics have become fairly standardized and often follow the ways the federal government gathers these data in the census and/or other surveys they conduct. Other common demographic measures identify the respondent's political party, political ideology, marital status, religious preference, church attendance, voter registration status, geographic place of residence, and number of children. Occasionally, the nature of a poll or other survey might cause specific other demographic questions to be asked, such as questions about military service, union membership, sexual orientation, type of employment, type of housing unit, and years lived in one's neighborhood.

...