Survey Research

Survey research is one of the most commonly used methodologies in the social sciences. Survey research refers to the set of methods used to gather data in a systematic way from a range of individuals, organizations, or other units of interest. Specific methods may include questionnaires (on paper or online), interviews (conducted by any method; e.g., individual interviews done face to face or via telephone), focus groups, or observation (e.g., structured observations of people using internet access stations at a public library). This entry focuses on methods most commonly associated with survey research: questionnaires (often used in quantitative research designs) and interviews (often used in qualitative research designs).

Many studies using more than one data collection method will include a survey method. For example, a quantitatively oriented questionnaire could be used to generate general understanding of a set of related questions, to identify interview questions for deeper qualitative investigation, and to identify possible interview participants. Alternatively, a questionnaire could be used to confirm the generalizability of results from a small interview study to a larger, more statistically representative sample. Timing of survey research is relevant to the data produced. For example a cross-section approach collects data at one point in time. Longitudinal survey research can sample data at different periods over a length of time to examine long-term trends.

Although some researchers believe survey research to be a wholly quantitative approach, this opinion is not universally shared. Data gathered from any survey method may be entirely quantitative, may be largely qualitative, or may be a mixture. For example, open-ended questions on a questionnaire or asked in an interview will produce text that may be analyzed qualitatively. Qualitative data gathered in survey methods tends to be in text form, such as narrative responses to open-ended questions posed in an interview or written responses to a “comments” item on a questionnaire. There are many ways in which open-ended text or commentary is analyzed, although content analysis (either quantitative or qualitative) would be a commonly applied analytic method. The value of this qualitative data collected during the course of a questionnaire can be particularly important to contextualize more quantitative responses and to add depth and richness to the data set. For example, a questionnaire seeking data about service delivery in a nonprofit setting may include space for open-ended comments where respondents can indicate why they responded in certain ways or provide details not otherwise captured by closed response items. Responses to open-ended questions can provide detail about perceptions, opinions, personal experiences, and deeply held beliefs. Of course, written responses assume basic literacy on the part of respondents, as well as motivation to take the time required to write comments.

Open-ended questions provide greater freedom to the researcher in terms of how to frame the question, as well as granting greater freedom to respondents in the ways they choose to answer. Open-ended questions are often used in qualitative research to explore an issue or concept, to obtain natural wording, to add variety to a questionnaire, to obtain exact numerical data, and to provide respondents with opportunities for self-expression or elaboration. Open-ended questions may challenge respondents because they are more demanding and time-consuming to answer; however, the data obtained are typically richer than that generated from closed questions. Researchers often find that responses to open-ended questions require greater effort to record, code, analyze, and interpret than is the case for responses to closed questions. However, closed questions are easier and faster for respondents to answer, and responses to closed questions are easier for researchers to code and analyze, and they provide consistent response categories. Closed questions have many disadvantages including eliciting responses where no knowledge or opinion actually exists, oversimplifying issues, and forcing answers into possibly unnatural categories. It is important to ensure that response categories to closed questions are inclusive of all reasonably possible responses and are nonoverlapping.

...