Real-time financial monitoring requires operators to interpret large volumes of multidimensional and continuously changing data under conditions that demand rapid, accurate, and repeatable judgment. Conventional tabular interfaces rely heavily on serial scanning and working-memory—intensive comparison, increasing cognitive load and limiting the timely detection of anomalies or the integration of multiple indicators. This study introduces Multiviz, a visualization system designed to support high-density financial analysis through a squarified treemap layout, perceptually separable multivariate encodings, and a rule-based agent layer that applies nonintrusive visual emphasis to contextually relevant changes. A controlled laboratory experiment was conducted with twelve finance-domain participants who completed six analytical tasks using both Multiviz and a baseline tabular interface. Quantitative results showed that Multiviz reduced task completion time from 241.5 ± 42.7 s to 113.1 ± 13.8 s and increased task accuracy from 7.75 ± 0.26% to 9.17 ± 0.27%. Mouse interactions decreased from 52.6 ± 5.4 to 34.9 ± 3.4, indicating reduced interaction overhead. Eye-tracking analysis further revealed shorter fixations (414 vs . 441 ms), reduced scanpath length, and more concentrated attention within semantically coherent regions, suggesting greater perceptual efficiency. Participants reported improved clarity and reduced search effort after brief familiarization with the system. These findings provide empirical evidence that integrating hierarchical layout, multivariate visual encoding, and rule-based perceptual cueing can improve performance and attentional focus in complex, data-rich financial analysis settings.
Göktürk et al. (2026) studied this question.