This paper reports research on several methodological biases, which can occur in all types of surveys, but does not specifically examine transport research. It discusses three important types of bias. Almost every survey asks people about past behaviours and experiences, but there are several ways of phrasing its questions and constructing its response alternatives. Biases can occur with response alternatives of any type for behavioural frequency questions, for which many other considerations need to be allowed for. 'Intensifiers', words like `extremely' and `very', can have a considerable impact when used with survey questions, but often produce no effects; it is recommended that they are not used. For several topics, researchers are interested in people's knowledge and how it relates to people's attitudes. The author's research shows that actually measuring people's level of knowledge of science can affect expressed interest and even alter what is meant by scientific terms.
Abstract