research-package · RP-ATLAS-COLOR-004
Research index
research-package
Documents grouped under research-package.
research-package · RP-ATLAS-TYPO-TRANSFER-001
Typography Cross-Layer Transfer and Adaptive Reading
This research package investigates when improvements to glyphs, fonts, spacing, size, width, and weight transfer to higher-level outcomes such as word recognition, continuous reading, comprehension, navigation, memory, and action. The research rejects a simple upward-transfer model. Typography changes can improve local perceptual capacity without improving reading, can preserve reading speed through increased compensatory effort, or can improve reading for some individuals while harming others. The strongest emerging theory is that transfer occurs when three conditions align: the manipulated typographic property affects the active bottleneck, the benefit survives interactions with layout and language, and the measured outcome is sensitive to the cost being reduced. Atlas should therefore model typography as a personalized adaptive control problem rather than search for one universally optimal typeface.
Start here · RP-CCE-0001
Clinical Communication Engineering Foundation Research Execution Package
research-package · RP-CLF-002
Semantic Durability Across Interface Contexts
This package defines the next foundational investigation for the component library: whether one durable semantic content model can support substantially different editorial, application, mobile, marketing, low-vision, and print presentations without inaccessible reordering, presentation-specific markup, or variant proliferation. Preliminary standards evidence supports semantic stability only when meaning and interaction remain stable. The package therefore treats stable markup as a bounded hypothesis, establishes competing models, measurable failure thresholds, evidence schemas, and handoff instructions for executing the study before production component APIs are designed.
research-package · RP-COMP-005
Visual Scene Construction, Predictive Processing, and Active Perception
Evidence supports a hybrid account in which rapid feedforward processing produces provisional scene and object hypotheses, recurrent and contextual interactions refine them, and active sampling obtains additional evidence. Prediction is important, but predictive coding is not established as the unique neural account.
research-package · RP-PROD-002
Operationalizing Product Legibility
Product Legibility can be measured, but not responsibly as a context-free aesthetic score. The evidence supports a task-, population-, and environment-specific profile built from six gates: action possibility, discoverability, interpretability, executability, feedback closure, and consequence comprehension. A single average is unsafe because severe failure in one critical dimension can be concealed by strength elsewhere. This REP replaces the proposed additive score with a gated measurement architecture, distinguishes diagnostic metrics from outcome metrics, and provides a repeatable Product Legibility Evaluation Protocol. It also finds that classic predictive laws such as Fitts' and Hick's are useful locally but cannot serve as universal product-quality laws without checking their assumptions.
research-package · RP-WC-0001