Causal Approach for Careless Responding
Date:
In this study, I use a causal graphical framework to characterize careless-response generation mechanisms and distinguish them from detection and handling decisions. Similar observed response patterns can arise from different mechanisms, with different implications for whether responses should be retained, removed, or adjusted. Through simulations, I examine how imperfect detection and handling strategies that do not match the data-generating process contribute to bias. This work clarifies why removing respondents flagged as careless does not necessarily eliminate bias and provides a framework for evaluating the assumptions underlying common handling practices. This study was selected to receive the Psychometric Society Travel Award sponsored by ACS Ventures.