Survey research: Elements of a “good” survey
January 2022 | Written by Judith J. Stephenson, SM
What is survey research?
Survey research is a critical component of patient-centered research involving gathering information systematically from individuals by asking them questions about a specific topic or topics and generalizing the results to a larger population. Surveys are especially useful in obtaining information from individuals that cannot be determined any other way (e.g., beliefs, knowledge, attitudes, opinions, or satisfaction).
After determining that a survey is the best method for obtaining answers to their research questions, researchers should consider the elements or building blocks of a “good” survey.
What makes a “good” survey?
A “good” survey is one that is designed to obtain clear, relevant responses to the survey questions in a consistent way from respondents such that estimates with known precision can be made for the population from which the respondents have been selected (that is, the results are generalizable).
Survey research is no different than other disciplines in that it has its own methods and language — there are numerous articles and textbooks on survey methods and doing survey research. One common approach that is frequently referred to in the literature is the Tailored Design Survey Methodology (TDSM), or “Dillman Method” developed by Dillman et al. (2014).
The Dillman Method*
According to Dillman, tailored design is a strategy that customizes survey procedures for each survey situation. Each survey is unique and survey procedures need to be modified to fit the situation based on factors including knowledge about the topic and survey sponsor, the survey population, and the available resources. His book provides guidelines (see the 2014 edition listed below) for designing and conducting mixed-mode surveys in which multiple methods of recruitment and survey administration are used to improve the accuracy of survey results.
Dillman’s general approach is based on three fundamental considerations: survey error reduction, survey procedure construction, and positive social exchange. These concepts are good to keep in mind when involved in studies involving survey research.
Three fundamental considerations
- Survey error reduction focuses on reducing four common types of survey errors (coverage, sampling, nonresponse, and measurement) that may affect the quality of the collected information and the generalizability of the survey results.Survey procedure construction emphasizes that there is more to survey research than creating the questionnaire.
- Survey procedure construction emphasizes that there is more to survey research than creating the questionnaire. Survey research also involves developing a set of survey procedures or processes that interact and work together to maximize the success of the survey — the whole is more than the sum of its parts.
- Positive social exchange leverages elements of survey research that can be enhanced through positive social exchange (e.g., how information about the survey is provided to the individual through things like the pre-notification letter, asking individuals for their help or advice, or giving tangible rewards) — it consists of all the things that are done to engage the individual to want to complete the survey. It assumes that there is a greater probability that an individual will complete the survey when the individual perceives that the expected rewards for responding to the survey outweighs the anticipated costs of responding.
The ability to conduct survey research in today’s world has become much more complex and utilizing mixed-mode survey designs are necessary for improving overall survey response rates and providing high-quality results.
*Dillman DA, Smyth JD, Christian LM: Internet, phone, mail, and mixed-mode surveys: The tailored design method (4th ed). Hoboken, NJ: John Wiley and Sons (2014).