State of the art, per sub-question

The research record and the practice record are kept apart, as in every topic (method). For figures, the best evidence is mostly about charts and interfaces.

Best evidence

Sub-question Record holder Measured how
Q11a Layout Purchase 1997; Huang 2008b experiments on comprehension of graph drawings; path tracing against crossing angle
Q11a Layout, the map of the field Burch 2021 survey of 239 papers of user studies of graph visualisation
Q11b Visual channels for quantities Cleveland 1984; Heer 2010 perception experiments, and their crowdsourced replication
Q11b Grouping Palmer 1992 laboratory demonstrations of grouping by common region
Q11c Appeal and performance Thielsch 2019; Schlamann 2026 meta-analyses, 25 and 31 studies, for interfaces
Q11c Embellishment Bateman 2010; Borgo 2012 controlled experiments on charts, description and recall
Q11d Reaching a figure without sight W3C 2023e; W3C 2023a a standard; not tested with readers
Q11d Reading by a language model Zhong 2024 500 random graphs, text against image, one model
Q11e Practice none no study of a style guide's rules was found

Most adopted

Practice Holder Sign of adoption
Ten rules for better figures Rougier 2014 the reference most practice pages cite
A stated message, direct labels, little decoration Urban 2025; Datawrapper 2022 newsroom and institute style guides
A vocabulary of chart types by purpose FT 2026 a newsroom's public guide
Diagrams drawn from source in documentation GitHub 2026 diagrams; Mermaid 2026c rendered in every Markdown file on GitHub