Identifying Optimizers, Extremists, and Indifferents: Latent Satisficing Patterns in Panel Surveys

Abstract

Data quality is known to be compromised when respondents cognitively shortcut the survey response process. This satisficing behavior leads to inaccurate and unreliable responses that are hard to compensate after data collection. Thus, detecting and understanding survey satisficing is crucial for developing and implementing effective preventive measures in longitudinal data collection contexts. We use repeated latent class analyses across three waves of a self-administered mixed-mode panel survey to identify patterns of satisficing. Moreover, we identify correlates and predictors of future satisficing. Results indicate that the same three classes (”Optimizers”, ”Indifferents”, and ”ExtreMists”) replicate over time. The identified classes differ in their socio-demographic composition and results vary across survey modes (paper versus web). Most importantly, the particular satisficing strategy in one wave is predictive of satisficing in the following wave on the individual level, suggesting potential for targeted interventions across panel waves.