Respondent-Related Error

Respondent-related error refers to error in a survey measure that is directly or indirectly attributable to the behaviors or characteristics of respondents; it is distinguished from error resulting from other survey components, such as questionnaires, interviewers, or modes of administration. However, respondents may interact with these other components of survey design in producing errors. It is useful to dichotomize respondent-related errors into those that arise from nonobservation or nonresponse (e.g. during efforts to obtain interviews) and those that result from observation or measurement (e.g. during the administration of the survey).

Errors in survey measures have two components: bias and variable errors. Bias results when responses provided in the survey differ from their true values, which are typically unknown and unmeasured, in a systematic way across repeated ...