Knowledge workers frequently face the challenge of balancing the desire for additional information against constraints of time, effort, and limited resources. While many systems aim to support decision making by providing users with ``just enough'' information through summarization or downsampling, having sufficient information does not necessarily align with users' subjective sense of having seen enough to act. This gap motivates the study of information satiety, defined as the subjective experience of reaching a stopping point in information seeking and exploration. This research investigates how information satiety emerges in visual decision-making contexts and how it varies across individuals. Experiment~1 is a completed empirical study that examines how individual differences influence stopping behavior and task performance in a controlled, interactive visualization setting. Focusing on personality traits (Big Five) and visual literacy (Mini-VLAT) as measurable individual differences, the study finds that personality traits are associated with stopping behavior, whereas visual literacy primarily predicts task accuracy, with no cross-effects observed. Additionally, the results reveal a systematic tendency for participants to stop far earlier than would be required for statistically reliable performance as task difficulty increases, suggesting a misalignment between subjective sufficiency and objective task demands. Building on these findings, this research proposes a follow-up mixed-methods study (Experiment~2) as future work. The proposed study is designed to qualitatively examine how individual differences and contextual knowledge shape participants' reasoning about when enough information has been obtained, using mental-model alignment as a key manipulation. Grounded in theories of information need and human sense-making, this work aims to advance understanding of information satiety as a human-centered boundary phenomenon and to inform the design of systems that better support informed stopping decisions across information-intensive domains.
Wenyuan Wang (Thu,) studied this question.