Survey Wording

A survey is a set of questions or statements to which participants give responses. A survey provides one of the best methods to obtain a large amount of data from participants. In this manner, survey data can provide a quantitative, qualitative, and/or numeric description of trends, attitudes, or opinions of a population by studying a sample of that population. The wording of the items of the survey is important because different words can generate different answers. When writing survey questions, researchers must take into consideration the stimulus function of the prompt. Survey questions should invoke a commonly understood concept and ask for a reaction or evaluation of that concept using commonly understood words and phrases. This entry provides some general considerations about survey wording. It then reveals some elements in survey design that researchers should avoid or use with caution when utilizing. Finally, ethical considerations regarding survey wording are reviewed.

Researchers designing a survey should carefully consider how participants from the sample population will interpret the words of the items. Among various survey challenges, poor wording is a main concern in threatening the quality of the survey. Issues may involve generational interpretation and [Page 1731]application of terms or whether separate samples respond differently to common expressions. A survey may have an intention of generating measurement of a specific outcome, but the wording of the survey may produce an outcome not sought by the researcher because the participants responded differently based on a misunderstanding of the survey wording. For example, there exist serious disagreements about the value and reaction to the terms “Latina/o” and “Hispanic” to describe the ethnicity, race, or nationality of a person. Participants may experience different reactions to the survey and thus may answer questions differently based on the interpretations that exist with a specific term.

The survey wording must also be clear and lead the participants to provide responses that match the intent of the survey. For example, suppose a survey asks participants about their preference for a political candidate. The question asked could be, “Which candidate do you support for the elected position?” Alternatively, the question could be worded, “For which candidate are you planning to vote?” The first question asks what candidate a participant favors, while the second question asks who the participant will vote for, and the participant could have two different answers to those questions. A person may support or agree most with a particular candidate, but that candidate may have no chance of winning the election, so the person may intend to vote for a different candidate that has a more realistic chance of winning the election. The distinction of the two questions is one of attitude about the suitability of a candidate versus the actions a person will take in regards to making a choice among candidates. While the two views are often related, the action and attitude may become inconsistent for very rational and understood reasons.