Patterns of error - response biases and heuristics
According to Bradburn, Sudman and Wansink (2004), four basic factors can lead to respondent error in self-reported survey measures: i) failures in memory (e.g. material may be forgotten or misremembered); ii) lack of appropriate motivation (e.g. some respondents may be motivated to present themselves in a positive light, or unmotivated respondents may not process questions fully); iii) failures in communication (e.g. the meaning of the question may be unclear or misunderstood); and iv) lack of knowledge (e.g. respondents may simply not know the answer to the question, but may attempt to answer it anyway). Respondent errors can be either caused or exacerbated by aspects of survey design and context - and different sources of error will often call for different solutions.1
Failures in memory, motivation, communication or knowledge are often associated with increased risk of response biases and the use of response heuristics. Response biases refer to particular patterns or distortions in how individuals or groups of individuals respond to questions; response heuristics refer to (often sub-conscious) rules or short-cuts that respondents may use in order to help them select their answers. Drawing on the classifications of Podsakoff et al. (2003), some examples of response biases and heuristics are described in Table 2.1. Where a respondent exhibits a repeated tendency towards a particular response bias or heuristic, this is referred to as a response style.