Chapter 8 Review Questions

Face-to-face surveys continue to be used for survey data collection. This data collection method provides the most versatile methodology because respondents can be presented with a mix of question formats and visual aids. These surveys are known as paper and pencil interviewing (PAPI) when responses are recorded by hand and as computer-assisted personal interviewing (CAPI) when responses are recorded onto a computer. The direct human contact can result in higher response rates and permit longer surveys, as respondents may be more inclined to complete a survey when the interviewer is directly in front of them. At the same time, face-to-face interviews may be most prone to the interviewer effect. These surveys are less frequently used than other forms because they are expensive and can be difficult to conduct in a security-conscious environment where people are uneasy about admitting strangers into their homes.

One of most common concerns in survey research is the representativeness of the resulting data, as researchers need to be aware that the results may be biased in some way. One issue is coverage bias, in which the survey data collection mode or some other factor excludes particular groups from the sampling frame. Early telephone survey research suffered from coverage bias because telephones were not present in many homes, and Internet surveys are limited in their ability to contact individuals without home Internet access. Another concern is non-response bias, a sampling error that occurs if the individuals who opt to participate in a survey are in some important way dissimilar to those who choose not to participate. Survey non-response rates are increasing over time, largely due to increasing refusal rates. This change raises concerns about the representativeness of survey results.

Secondary data are data that are not directly collected by the researcher. Secondary data present both advantages and challenges. A key advantage is efficiency and cost-savings. By using secondary data, the researcher saves data collection time: rather than spending months designing and implementing a study, they can access data almost immediately. Using secondary data sets also saves the researcher the (often considerable) costs associated with data collection. Many data sources are freely available from public websites. Another advantage can lie with research design. Some secondary data sources are longitudinal. Longitudinal data allow researchers to consider change over time and to clarify if certain events preceded others, thereby enabling researchers to better assess causality. There are some limitations to using secondary data sets as well. One chief drawback is that secondary data sets often include some measures of interest to the researcher but not all of what they would have included in an original design. Similarly, the questions included in the survey may be worded differently than the researcher would have preferred.

Question order is important for survey research. Researchers commonly include a few innocuous and non-threatening questions at the outset of a survey. These questions occur early in the order because they are often the basis for the survey company’s overall assessment of voting intentions. It is generally desirable to obtain information on such questions before asking respondents to assess factors that might influence their voting decision or opinion of the leaders. If there were a series of questions about the parties’ stands on issues and a respondent indicated a preference for the position of party Y, it could create mental stress for the person to indicate an intention to support party X. Once the vote-intention questions have been asked, questions about issue influences and perceptions, the party leaders or their personal characteristics, and general deep-rooted values and beliefs typically follow. These questions can be either open or closed depending on the needs of the researcher and the timeliness of their use. Once the full set of attitudinal and behavioural questions are asked, the final section of a survey asks about respondents’ personal demographic characteristics. These questions generally are placed at the end of a questionnaire because some respondents do not like to answer them and because asking them too soon could lead people to wonder whether they will really retain their anonymity. Further, such questions are considerably less interesting, and leading with these questions could result in greater termination (and thus non-response) rates.

Aggregate data is grouped data that is applies to a particular place or group of people. Many official statistics are available as aggregate data. Because aggregate data provide group-level (as opposed to individual-level) information, researchers must be careful to ensure that their conclusions do not imply individual-level behaviours. Microdata, the individual level data files, is less likely to be available from official sources; it is highly restricted because of strict confidentiality rules. Metadata is the technical documentation associated with the data file. It is critical that researchers review the metadata to avoid unintentionally misusing or misrepresenting secondary data.