Blackwell Reference Online

Extract

This term, or its complement – nonresponse rates – refers to the level of cooperation of respondents in a → survey . It is often used as an overall measure of the quality of a data collection, although researchers agree that this is an incomplete and inadequate indicator of quality. Since surveys and polls are a common form of data collection in communication research, understanding the impact of response rates on research results is important. The topic is especially salient as there has been a decline in response rates across a wide range of survey types for the last several decades, raising basic questions about the general quality of – or the biases inherent in – data collected through surveys (→ Interview, Quantitative ; Public Opinion Polling ; Election Polls and Forecasts ). There would not be as much concern about the level of nonresponse if information to be gathered in a survey were lost at random. However, the general suspicion is that the nonresponse rate represents a potentially serious source of bias in a survey because the loss of information is in fact not random. There may be specific reasons why some people or groups of respondents refuse to participate in a survey or to answer particular questions. In such cases, nonresponse can introduce a bias in the findings because these individuals have certain characteristics that may be related to the central focus ... log in or subscribe to read full text