Purpose This paper aims to examine the cognitive processes involved in answering survey questions. It also briefly discusses how the cognitive viewpoint has been challenged by other approaches (such as conversational analysis). Design/methodology/approach The paper reviews the major components of the response process and summarizes work examining how each of these components can contribute to measurement errors in surveys. Findings The Cognitive Aspects of Survey Methodology (CASM) model of the survey response process is still generating useful research, but both the satisficing model and the conversational approach provide useful supplements, emphasizing motivational and social sources of error neglected in the CASM approach. Originality/value The paper provides an introduction to the cognitive processes underlying survey responses and how these processes can explain why survey responses may be inaccurate.