Confidence tiers have been paired with multiple choice items across different fields since the early twentieth century and have seen widespread adoption in discipline-based education research fields seeking to evaluate aspects of self-regulated learning. The design of two-tiered confidence judgments impacts interpretability and perception of their utility, so meaningful engagement hinges on intentional design for specific constructs. This study uses cognitive interviews exploring students' interpretations of confidence tier components to identify design features which elicit meaningful variation in self-reflection. The evidence supports two prevailing motivations for using confidence tiers--prompting metacognition and measuring the strength of learners' alternate conceptions. The challenges and strategies students encounter while engaging with confidence tiers serve as the foundation to improve the validity of collected confidence data. Recommendations are presented to improve the clarity and utility of confidence tiers to provide meaningful evaluation of target constructs.