ESRA 2027 sessions by theme

Back to Overview of Sessions

Session Details

In an ideal world, all survey respondents would be real, honest humans who fully go through the cognitive response process when answering our questions, including carefully comprehending questions, retrieving relevant information from memory, formulating an answer based on informed judgments, and reporting an accurate answer using the provided responses (Tourangeau et al., 2000; Tourangeau & Bradburn, 2010). However, this is not always the case, as respondents often exhibit satisficing—taking cognitive shortcuts to reduce the effort needed to provide an optimal answer (Krosnick, 1991).

Decades of research have focused on identifying and removing respondents who provide suboptimal responses; however, as online surveys have skyrocketed and bots and AI are increasingly infiltrating surveys, researchers have had to develop new methods to identify both suboptimal and fraudulent respondents. While many methods have been developed to identify different types of suboptimal and fraudulent respondents, these measures often add time and effort to the survey. In addition, removing respondents may introduce other biases into the data if certain types of respondents are more prone to these behaviors. As such, it is important to examine how this data cleaning affects substantive results to understand when, how, and whether to clean respondents from the data.

This session seeks papers on how various forms of data cleaning for suboptimal or fraudulent respondent behavior impact substantive results relevant to survey researchers, such as accuracy relative to high-quality benchmarks, statistical power of hypothesis tests, relationships between variables, and the effects of experimental manipulations. Data cleaning for suboptimal or fraudulent respondent behavior may be based on indicators of speeding, item non-response, grid non-differentiation or straightlining, response styles, consistency checks, attention checks or compliance traps, bot or AI identification questions, or other cleaning indicators.