Response

Response to the Request to Participate

At the time a person is notified of having been sampled to participate in a survey, he or she may decide to participate or not. This response, whether or not to participate, affects the response rate of the survey, which essentially is the proportion of eligible respondents who agree to participate. Response rates may vary as the result of a number of factors, including the population being studied, the procedures used for sampling, the mode of the survey (e.g. telephone, in person, mail, or Web), questionnaire length, field period length, number of contact attempts, the topic of the survey, and whether procedures such as advance letters, refusal conversions, and incentives are used.

Although no specifie response rates are generally required for conducting a survey, researchers traditionally strive for high response rates. Low response rates may lead to nonresponse bias when certain groups of respondents are more likely to participate than others, which creates the phenomenon of under- or overrepre-sentation of the certain attributes that the survey is striving to measure. As such, it is in a researcher's interest to encourage an affirmative response to the request to participate from as many respondents as possible that are sampled.

Response to Survey Questions

When someone is asked a survey question, the effort that goes into determining and then providing an answer (the response) can vary considerably from respondent to respondent. Thus, respondent-related error is of considerable consequence to survey researchers. Such error may be introduced into survey responses to individual survey questions in a number of ways, such as when respondents answer survey questions based on what the survey designers expect.

This may occur because respondents want to please researchers or because of leading questions. Response error may also be introduced if respondents misinterpret or do not understand a question. The format of survey questions (e.g. whether they are open- or closed-ended, the number and specific response choices provided) influences respondents' comprehension of a survey question. Response error can be minimized if researchers consciously implement strategies while designing the instrument, such as avoiding leading questions and wording questions precisely. The order of questions in surveys can also contribute to response error, although there are also some general guidelines for minimizing error from this source (e.g. easy questions should be placed before hard ones).

Respondents themselves can also be a source of response error because respondents may not be willing or able to answer the questions with correct information. Thus, response error may be introduced if respondents satisfice or engage in socially desirable responding. These issues can also be somewhat alleviated through optimal questionnaire design.