Designing Annotations in Visualization: Considerations from Visualization Practitioners and Educators Volume: 45 Issue: 3

Md Dilshadur Rahman 1 Devin Lange 2 Ghulam Jilani Quadri 3 Paul Rosen 1 1 Scientific Computing and Imaging Institute, University of Utah, USA 2 Harvard Medical School, USA 3 University of Oklahoma, USA · 2026

Read Designing Annotations in Visualization: Considerations from Visualization Practitioners and Educators Volume: 45 Issue: 3 by Md Dilshadur Rahman 1 Devin Lange 2 Ghulam Jilani Quadri 3 Paul Rosen 1 1 Scientific Computing and Imaging Institute, University of Utah, USA 2 Harvard Medical School, USA 3 University of Oklahoma, USA (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-07691.txt.

Introduction

Rahman and three colleagues examine how annotations are made in practice and how educators reflect on them. Prior work has covered the visual forms they take, but not the choices behind them. A two-phase study interviews ten pros and seven visual teachers, showing their rules for making notes and offering views on clarity and viewer agency. The studies give an account of note design knowledge, highlighting factors and local judgments. Making this know-how open adds to form-focused studies and points to better tool and guide support.

Catalog record

CREATORMd Dilshadur Rahman 1 Devin Lange 2 Ghulam Jilani Quadri 3 Paul Rosen 1 1 Scientific Computing and Imaging Institute, University of Utah, USA 2 Harvard Medical School, USA 3 University of Oklahoma, USA
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66454
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.07691

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Designing Annotations in Visualization: Considerations from Visualization Practitioners and Educators Volume: 45 Issue: 3 Md Dilshadur Rahman 1 Devin Lange 2 Ghulam Jilani Quadri 3 Paul Rosen 1 1 Scientific Computing and Imaging Institute, University of Utah, USA 2 Harvard Medical School, USA 3 University of Oklahoma, USA Abstract Annotation is a central mechanism in visualization design that enables people to communicate key insights. Prior research has provided essential accounts of the visual forms annotations take, but less attention has been paid to the decisions behind them. This paper examines how annotations are designed in practice and how educators reflect on those practices. We conducted a two-phase qualitative study: interviews with ten practitioners from diverse backgrounds revealed the heuristics they draw on when creating annotations, and interviews with seven visualization educators offered complementary perspectives situated within broader concerns of clarity, guidance, and viewer agency. These studies provide a systematic account of annotation design knowledge in professional settings, highlighting the considerations, trade-offs, and contextual judgments that shape the use of annotations. By making this tacit expertise explicit, our work complements prior form-focused studies, strengthens understanding of annotation as a design activity, and points to opportunities for improved tool and guideline support. \SpecialIssuePaper \CGFccby graphicx † † year: 2026 † † year: 2026 † † editors: N. Andrienko, M. Chen, and B. Wang † † subject: Eurographics Conference on Visualization CGF Vol No † † orcid: 0009-0008-5467-615X † † orcid: 0000-0002-3467-0294 † † orcid: 0000-0002-8054-5048 † † orcid: 0000-0002-0873-9518 † † teaser: The sketches illustrate six annotation design considerations derived from interviews with practitioners (P1–P10) and educators (E1–E7). 1 Audience: decide what context must appear on the chart so it remains interpretable when reused without surrounding text. 2 Hierarchy: make the primary annotation dominant and secondary notes visually subordinate. 3 Placement: place text next to its target when possible, use short connectors when proximity is infeasible, and use a key or legend only when direct attachment would clutter the view. 4 Association: use consistent cues, such as color and styling, to link annotation text to its target data element. 5 Cohesion: decide whether annotations blend with the chart or stand apart as a distinct layer. 6 Amount: keep only the annotations necessary to support the main message, removing those that clutter the view. We provide a companion gallery that maps public professional examples from a published dataset [ RQR23 ] to these considerations: vis-annotations.github.io/annotation-design . 1 Introduction Amanda Cox, former Graphics Editor of The New York Times and a leading figure in data journalism, once remarked, “the annotation layer is the most important thing we do… otherwise it’s a case of here it is, you go figure it out” [ Kir12 , RBea17 ] , underscoring that annotations are central to how visualizations convey meaning. We use annotations to refer to textual or graphical elements that form an author-supplied communicative layer over the underlying encodings, directing attention, highlighting salient features, providing context, and guiding interpretation [ RQD∗25 , Mun14 ] . Annotations can improve comprehension, recall, and engagement [ BVB∗13 , ALX∗21 , ZOM19 , BWH∗23 , MDF12 , BMSM20 ] , support mental model construction of data [ CR00 ] , and externalize analytic findings for collaboration and narrative communication [ KS14 , SEAB∗22 , RHRea19 , CAB∗11 , MT14 , SH10 , HD11 , SH14 , GZL∗20 ] . Visualization guidance discusses annotation-related design choices such as labeling, emphasis, and narrative framing [ Mun14 , War04 , Tuf01 , Cai16 , SOK∗15 , SH10 ] , and research has examined annotations via professional visualization analyses [ RQR23 , SH10 ] , authoring tools and grammars [ RBea17 , HDA13 , RZMR25 , CWS∗25 , LSB19 , SBea15 ] , artifact analyses [ RQD∗25 , SAea25 ] , and audience effects in chart and text integration [ SSC∗22 , BBK∗15 ] . These efforts clarify common forms of annotation, document audience effects in specific settings, and propose systems that scaffold annotation authoring. However, they provide limited empirical visibility into how visualization authors make annotation decisions in context. In practice, visualization authors must decide what evidence warrants an annotation, what context or caveats must be made explicit in the visualization itself, and how to place and style annotations while managing attention and annotation density. These choices are consequential: annotations are often the primary channel through which authors communicate intended interpretation alongside the visual encodings. Without an account of this decision logic, annotation guidance can remain abstract: visualization authors may learn principles, but still lack support for the decisions and trade-offs required to produce effective annotations within the constraints of a given visualization, audience, tools, and medium. We use the term practitioner to denote professionals who regularly design, produce, or communicate with data visualizations, such as data journalists, analysts, visualization designers, and consultants. Their annotation choices shape how audiences interpret charts, yet the reasoning behind these choices remains largely undocumented. Design scholarship argues that expert work relies on tacit, situated knowledge, and that making such knowledge explicit can surface the considerations and trade-offs that organize practice and support practice-grounded tool and pedagogy design [ Pol09 , Sch17 , Suc87 ] . Prior practice-grounded studies have characterized visualization design workflows and judgment in professional settings [ Par21 , PGBC20 , PS20 , PSP21 , AZL∗18 , BDea16 , BDea14 , HLea20 , Bea25 , PR25 ] , but comparable accounts of annotation-specific decision-making remain underexplored. To address this gap, we conducted a two-phase study that brings practitioners’ accounts of annotation design into dialogue with educators’ reflections on those accounts. We include educators because they regularly translate annotation guidance into critique and instruction, and can help surface boundary conditions and risks that practitioners may not articulate explicitly. In Phase 1 , we interviewed ten practitioners from diverse domains who regularly create annotated visualizations and elicited the considerations that guide their everyday annotation decisions. In Phase 2 , we interviewed seven visualization educators and researchers , presented the practitioner-derived considerations, and asked them to reflect on boundary conditions, risks, and teaching-oriented framings. The aim was not to validate practitioners or privilege one group over the other, but to use structured peer debriefing to surface key strategies, tensions, and trade-offs in annotation design [ Spa98 ] . We make three contributions. First, we provide a practice-grounded account of annotation design in professional visualization work, distilled into six design considerations that capture recurring annotation-specific decision points and trade-offs with examples at vis-annotations.github.io/annotation-design . Second, we juxtapose practitioner and educator perspectives, showing where they align, where they diverge, and which boundary conditions and tensions these differences reveal for annotation practice. Third, we outline implications for visualization design, tool support, and teaching, treating these considerations as a shared vocabulary for design critique, tool development, and future empirical work on annotated visualizations. Throughout, we present these considerations as heuristics for reflection rather than universal prescriptions. 2 Related Work Annotations in Visualizations. Annotations augment a visualization with author-supplied text and graphics that reference marks, regions, or coordinate space and shape how viewers connect evidence to interpretation [ Mun14 , RQD∗25 ] . Prior work has formalized annotation design spaces by characterizing forms, functions, and referential scope, including distinctions between annotations that introduce

Read the complete text