Task / Domain

Lesson 14: Tasks and Domains

Understanding tasks and domains, drawing information design to user centered design.

Objective

Deepen understanding of how to approach data visualization design through the lens of user tasks and domain contexts, introducing Tamara Munzner's What/Why/How framework as an optional structured approach.

Outline

This lecture expands on the idea of domains and tasks with examples before looking at the What / Why / How framework from Munzner. It provides more information on the second perspective introduced in Lecture 1, focusing on visualization as task alongside representation, narrative, and dialogue perspectives.

Understanding domains and contexts

Understanding the user's context is essential for effective visualization design.

• Who is the user and what is their expertise level?
• What social context is that user in (professional, academic, public)?
• What are the concepts the user is interacting with and their mental models?
• What prior knowledge or understanding might the user be interacting with?
• How are the data produced (collection methods, measurement instruments, sampling)?

Methods for understanding users

Three fundamental approaches to understanding users and their needs.

• Talk to them through user interviews and conversations.
• Ask them to share their work with you through contextual inquiry.
• Share prototypes with them for iterative feedback and testing.

There are structured appraoches but these act as general principles.

Understanding tasks and questions

What users need to accomplish is critical for design decisions.

• What questions are the users trying to answer or actions they need to take?
• What will the user do with that answer or action?
• Why is the task important (business value, scientific contribution, policy implications)?
• What information is needed to answer that question?
• Understanding tasks tells you what to encode and with what priority.

The What / Why / How framework

This framework from Tamara Munzner provides a structured approach to visualization design through three iterative questions operating in a nested manner.

• What (data abstraction): datasets, attributes, dataset availability.
• Why (task abstraction): actions (analyze, search, query) and targets (trends, outliers, features).
• How (visual encoding and interaction): encode, map, manipulate, facet, reduce.

Note that this can be a useful framework but is optional in your projects for the course.

Data abstraction

Understanding what we are visualizing through data structure and types.

• Data types: items, attributes, links, positions, grids.
• Dataset types: tables, networks and trees, fields, geometry (spatial), clusters, sets, lists.
• Attribute types: categorical, ordered (ordinal, quantitative).
• Ordering direction: sequential, diverging, cyclic.
• Dataset availability: static or dynamic.

Task abstraction

Understanding why users examine data through actions and targets.

• Analyze actions: consume (discover, present, enjoy) or produce (annotate, record, derive).
• Search understanding: lookup, browse, locate, explore based on what is known.
• Query actions: identify, compare, summarize.
• Targets for all data: trends, outliers, features.
• Targets for attributes: distribution, extremes, dependency, correlation, similarity.
• Targets for networks and spatial: topology, paths, shape.

Visual encoding and interaction

How the visualization helps achieve goals through arrangement and manipulation.

• Encode and arrange: express, separate, order, align, use marks effectively.
• Map from categorical and ordered attributes: color (hue, saturation, luminance), size, angle, curvature, shape, motion.
• Manipulate: change viewpoint or parameters, select items or regions, navigate through data space.
• Facet: juxtapose (side-by-side), partition (divide), superimpose (overlay).
• Reduce: filter unwanted data, aggregate and summarize, embed into other contexts.

Case study: NASA spacecraft operations

Rachel Binx's Vortex project revealed periodicity of events users had never seen visually before.

• Domain: NASA spacecraft operators monitoring real-time spacecraft with prior work of manual log review.
• Tasks: monitor spacecraft status, identify anomalies, understand event patterns, track periodic events.
• Design: visual representation of temporal patterns supporting identification of rhythms through multiple coordinated views.

Case study: Jigsaw visual analytics

An intelligence analysis tool for making sense of large document collections.

• Domain: intelligence analysts and researchers in investigative analysis contexts.
• Tasks: explore connections between entities, identify patterns across documents, compare different data views, generate and test hypotheses.
• Design: multiple coordinated views (list, graph, scatter plot, text), interactive linking, iterative exploration, focus on user journey through questions.

Take Aways

Effective visualization design requires understanding both the user's domain context and their specific tasks, with the What/Why/How framework offering an optional structured approach to design decisions.

• Talk to users, share work with them, and prototype iteratively to understand domain and context.
• Understanding tasks directly informs what to encode and with what priority.
• The What/Why/How framework provides a nested, iterative approach through data abstraction, task abstraction, and visual encoding.
• Apply these concepts to familiar examples through analysis of domain, context, and tasks.

Note

This outline was expanded after the video was recorded and may include additional details.

Citations

[1] T. Munzner, "Visualization Analysis & Design," University of British Columbia, 2021. Available: https://www.youtube.com/watch?v=pHljd-cgICY

[2] T. Romanowski, "Visualisation Analysis & Design by Tamara Munzner or the What-Why-How of data viz," Datarocks, 2023. Available: https://www.datarocks.co.nz/post/data-viz-bookshelf_visualization-analysis-design-tamara-munzner

[3] J. Stasko, C. Gorg, Z. Liu, and K. Singhal, "Jigsaw – Visual Analytics," Georgia Institute of Technology, 2009. Available at: https://youtu.be/2CMw4i9DiaM?feature=shared

[4] R. Binx, "Designing for Realtime Spacecraft Operations," OpenVisConf, 2016. Available at: https://youtu.be/HuYKhSHcRSQ?feature=shared

License

This lesson is part of Interactive Data Science and Visualization and is released under a CC-BY-NC 4.0 license.