Researching Human Subjects: Surveys and Secondary Data

This chapter examines quantitative forms of data collection from human subjects. Survey research, is one of the most popular methods of generating primary data to study political phenomena. Survey research includes telephone, Internet (or web-based) mail, and (less frequently) face-to-face surveys. Omnibus surveys, which allow researchers to add a few questions to a larger survey being conducted by a commercial firm, are another option. Each approach has its strengths and limitations, and the choice of one method over another can be determined by budgetary concerns, the research questions that need to be addressed, security concerns, and the resources available to the researcher.

Key concerns across all survey forms are also discussed. One major issue is the representativeness of the resulting data, which is affected by coverage bias and non-response bias. Cost is another important factor. Longitudinal studies, for example, can be expensive to conduct because contact with respondents must be maintained or renewed; the longer the time interval between survey waves, the more complicated the task of tracking respondents becomes.

In addition to collecting their own survey data, researchers rely on secondary data, such as official statistics or surveys fielded by other scholars. Secondary data can save time and money, and also allows researchers to access data they might not have been able to gather on their own. However, researchers must carefully review the metadata, including the research design used to collect the data, to ensure that data is robust and was collected ethically.

The chapter also provides practical suggestions on how to draw good quality measures from survey responses by highlighting the need to think about question type, question wording, and question ordering.