Response bias. A response bias is a tendency to respond on some other basis than the content of the items (Xxxxxxx, X. X., 1991). Research after the effect of response bias has been investigated since the forties, especially its (negative) effect on validity. Response biases distort relations and elevate variance in data (XxXxxxxx, X., & Xxxxxx, X. X., 1965). Although these biases prohibit a response reflecting a true opinion, it does reflect characteristics of the respondents and appears to be a consistent component (Xxxxxxxxxxx, X., & Xxxxxxx, S., 1959). A short list of important categories will be listed in the following paragraphs. Respondents can adjust their answers according to a certain image they want to reflect on people. This is called a response set (Xxxxxx, X., Böhnke, X. X., & Xxxxx, X., 2016). They could for example adjust their response to what they believe to be the socially desirable answer. This is called socially desirable responding and is one of the bigger problems within surveys (Xxxxxxx, P. M., Xxxxxxx, X. X., & Xxxxxxx, S., 1971). If positive wording is used, and only then, this bias can be synonymous with ARS (Xxxxx, X. X., 2004). The opposite can happen just as well. Socially desirability bias can overpower the volition to acquiesce (Xxxx, X. X., & Xxxxxxxx, X., 1984). Xxxxxxxxx, X. X. (1964) provided a tangible example with the evaluation of the statement ‘the trouble with most white people is that they think they are better than other people’. When there was a white interviewer present, there was 21% less ARS with all other factors held constant. The reduction in ARS was allotted to social desirability bias. Simulation and dissimulation are other examples of possible response sets. They entail the overreporting or underreporting of symptoms (Xxxxxx, X., Böhnke, J. R., & Xxxxx, X., 2016). Another category of response biases is response order effects. This is the tendency to indicate answers based on its placement. Some people have the tendency to choose the first response presented to them, called primacy effect. Other people are more inclined to choose the last presented item because of recency effect (Xx Xxxxxxxxxx, P. & Xxx Xxxxxxx, P., 1999). A third category of response bias is misresponse. This is the tendency to answer the same way to two opposite questions. This can find his origin in acquiescence, in item verification difficulty and respondent inattention (Xxxxx, S. D., Xxxxxxxx, D., & Xxxxxxxx, X. X., 2008). Last but not least, respo...
Response bias. It is possible that surveyed and interviewed individuals may exhibit a ―social desirability bias‖ by reporting answers that they perceived to be what the researcher wanted to hear or what were otherwise deemed appropriate by society and culture. Attempts to reduce the effects of response bias to gain the approval of cultural outsiders included the use of trained Liberian enumerators and research assistants who were familiar with cultural values and could speak to respondents in a trust-building manner. All survey questions were also couched in culturally relevant terminology in order to ensure understandability and accurate responses. Whenever possible, enumerators and focus group facilitators were conducted by an individual of the same sex as the respondent in order to facilitate trust and confidentiality of responses. Given that focus group participants were often selected through gatekeepers in the community, including community leadership and women’s group leaders, selection bias may have an effect on the quality and diversity of responses provided in focus groups. Individuals hand-selected by community leaders were assumed to be more active in the community and may be more informed about social issues or have higher levels of formal education than the average community member. In order to confirm or negate the effects of the selection bias present in focus group selection, strong attempts to ensure representativeness in the quantitative survey were made a chief priority of the overall research design.
