This study explores the alignment between designers' communicative goals and the high-level patterns that viewers naturally extract from visualizations. The researchers conducted a qualitative study involving 24 participants who described three types of visualizations—line graphs, bar graphs, and scatterplots—using natural language and think-aloud protocols. The study found that: 1. **Intent vs. Comprehension**: Only 41% of responses completely matched the designers' stated objectives, indicating that the intended communication goals often do not align with the patterns people intuitively comprehend. 2. **Cued Tasks vs. High-Level Comprehension**: Results from traditional cued tasks may not predict the knowledge people build with a graph, suggesting that low-level tasks alone are insufficient to address the real communication goals of visualizations. 3. **Chart Type Limitations**: Chart type alone is not sufficient to predict the information people extract from a visualization, as data type and graph complexity also play significant roles. The study highlights the need for more comprehensive guidelines that consider both low-level statistical studies and high-level comprehension to optimize visualization design and communication effectiveness.This study explores the alignment between designers' communicative goals and the high-level patterns that viewers naturally extract from visualizations. The researchers conducted a qualitative study involving 24 participants who described three types of visualizations—line graphs, bar graphs, and scatterplots—using natural language and think-aloud protocols. The study found that: 1. **Intent vs. Comprehension**: Only 41% of responses completely matched the designers' stated objectives, indicating that the intended communication goals often do not align with the patterns people intuitively comprehend. 2. **Cued Tasks vs. High-Level Comprehension**: Results from traditional cued tasks may not predict the knowledge people build with a graph, suggesting that low-level tasks alone are insufficient to address the real communication goals of visualizations. 3. **Chart Type Limitations**: Chart type alone is not sufficient to predict the information people extract from a visualization, as data type and graph complexity also play significant roles. The study highlights the need for more comprehensive guidelines that consider both low-level statistical studies and high-level comprehension to optimize visualization design and communication effectiveness.