We will look at how we can consider visualizations in terms of building blocks including data and task abstractions and visual encodings. We will see how this allows designing and analyzing visualizations. We will practice design process (including critique). Students will begin to design and analyze visualizations (pen-and-paper).

In this module, we will consider the building blocks of visualization. This includes both the building blocks for thinking about visualizations as well as the building blocks for making visualizations.

This is a lot to pack into a module - so we’ll go through them quickly, and return to them as the semester continues. There’s too much to read about everything, so we’ll be selective.

The building blocks of how we’ll think about visualization that we’ll need for class:

• Critique - This is the methodological tool we use to learn from examples. Getting good at it is a valuable skill for class and beyond. We’ll read some introductory material and practice in class.
• Task Abstraction - This is how we describe what we are trying to do with a visualization. I’ll introduce it in lecture, and add to it with reading.
• Data Abstraction - This is how we describe the data we are trying to visualize. A lot of this should be familiar if you are a programmer, mathematician, … There is some vocabulary building based on what is most useful for doing vis. A little bit of lecture, a little bit of readings.

The building blocks that we use to make visualizations:

• Encodings - how we map data to visual elements.
• Layout - how we use position - this is a special type of encoding, but it is so special that it deserves to be treated separately.
• Data Transformations - we often transform the data to make things easier to see in it. It’s important, but there is nothing to read about it. We’ll talk about it a bit.
• Interaction - providing the viewer with control over the visualization is a useful tool, but one we’ll explore later in the semester.

It’s a lot to learn - and everything wants to be first. I am trying a new strategy to cram it into less time, with less required reading, and less lecturing.

In this module, we’ll also start making visualizations. In week 2, Design Warmup 1: Ask Questions, Sketch Answers (due Fri, Sep 25) will have you create visualizations by sketching. In week 3, Design Exercise 1: Sketches and Questions (due Fri, Oct 02) will have you sketch ideas to solve some visualization problems.

Week 1-1 Sep 14-18 (Module 1)

Schedule-wise, week 1 has you focusing on the key readings. Since the week 1 readings aren’t due until the end of the week (with Reading Survey 1: Building Blocks (due Fri, Sep 18)), we won’t assume people have read things for this week’s lectures: instead, we will use the lectures to discuss things that aren’t covered well in the readings.

We’ll use some ICAs to make you appreciate the value of the topics. In addition, you’ll also do Seek and Find 1: Data, Tasks, Encodings (due Fri, Sep 18) to see how the concepts we’re learning play out in a real-world example.

Week 1-1 Readings

Read the week 1 readings before you take Reading Survey 1: Building Blocks (due Fri, Sep 18).

This week, you need to get up to speed on a bunch of the different building blocks we will need for class. There are alot of different pieces to get all at once. And each piece has some amount of reading to go with it.

The main thing is to read a little bit about each of the topics. I tried to make sufficient summaries for each topic. And then you can read more if you’re interested.

My intent was to write tutorials about each of the topics to give you a quick summary. And then you could read more as you had time. However, I didn’t get to actually write the tutorial on encodings - I’ll give you a “Claude-authored” tutorial, but probably you should read the Munzner chapters.

• backbone Michael Gleicher. Tutorial 2: How to Think about Visualization: Building Blocks. Vis Snacks Tutorial 2. (url)
• backbone Michael Gleicher. Tutorial 4: Critique. Vis Snacks Tutorial 4. (url) - read my tutorial to be ready for class. Reading more is optional.
• backbone Michael Gleicher. Data Abstraction Cheat Sheet. Vis Snacks Tutorial. (url) - my summary of the key things. Read it first. Read What: Data Abstraction if there is anything you need more clarity on.
• backbone Michael Gleicher. Task Abstraction: A Crash Course. Vis Snacks Tutorial. (url) - my summary of the key things. Read it, and read more if you are interested - it’s full of pointers.
• backbone Claude (Mike can't take credit). Encodings: Building Blocks for Visualizations. Vis Snacks Tutorial. (url) - this is Claude’s summary of encodings. It’s decent, but I think you should read the two munzner chapters to get the key ideas.
  • core Tamara Munzner. Marks and Channels. Chapter 5 from Munzner's Visualization Analysis & Design. (Canvas File) (video) (UW Library) - Munzner covers most of encodings in this chapter.
  • core Tamara Munzner. Arrange Tables. Chapter 7 from Munzner's Visualization Analysis & Design. (Canvas File) (UW Library) - Munzner divides layout (position encodings) from all other encodings, so this is a separate chapter

Week 1-1 Assignments

Note: you will need the readings for the seek and find.

• Reading Survey 1: Building Blocks (due Fri, Sep 18)
• Seek and Find 1: Data, Tasks, Encodings (due Fri, Sep 18)

Week 2 Sep 21-25 (Module 1)

In week 2, we’ll make use of the building blocks. In class, we’ll focus on making the concepts more concrete through examples, and do ICAs to practice putting the building blocks together. Design Warmup 1: Ask Questions, Sketch Answers (due Fri, Sep 25) will have you create visualizations by sketching to give you some practice.

Week 1-2 Readings

Read the week 2 readings before you take Content Survey 1: Building Blocks (due Fri, Sep 25). They are experiential readings.

For this week, we have 3 experiential readings (sorry, I had to add an extra - but it’s quite light).

One reading is an essay to add to our readings on critique. It’s sortof halfway between historical and modern. It isn’t a paper: it’s an essay. Some context: when they wrote this, the authors were already becoming famous for being the bridge between the academic research community and the practitioner community.

• historic Fernanda Viegas and Martin Wattenberg. Design and Redesign. Medium Posting. (Canvas File) (url) (Summary) - This essay is really valuable. And it is a case where the document itself is historically interesting, and their examples are worth looking at. This is valuable for seeing how far we’ve come at bringing design to visualization.

The “regular” experiential readings are two papers about Graphical Perception - the “science” of trying to understand encodings empirically.

The key context: the whole field of graphical perception started with a paper by Cleveland and McGill in 1984. There were early perceptual science papers, but this is the first time “visualization researchers” (in scare quotes since in 1984 they wouldn’t have identified themselves this way) tried to do experiments to inform chart design choices.

I am not requiring you to read the original 1984 paper (it is optional). Instead, I want you to read a 2010 paper that reproduces and extends the results, and a more current paper that challenges the paradigm.

• historic Jeffrey Heer, Michael Bostock. Crowdsourcing Graphical Perception: Using Mechanical Turk to Assess Visualization Design. (doi) (web pdf) (url)
• modern Caitlyn M. McColeman, Fumeng Yang, Steven Franconeri, Timothy F. Brady. Rethinking the Ranks of Visual Channels. IEEE Transactions on Visualization and Computer Graphics ( Volume: 28, Issue: 1, January 2022). (doi) (web pdf) (url)

The first paper was one of the earliest examples of using crowd-sources participants for visualization studies - something which is now totally standard practice. Many of us were starting to try it at the time (Detection of Image Stretching was our paper that came out around the same time). But Heer and Bostock really got it in widespread attention.

The second paper is notable not only because of its conclusion and interesting method but also for its statistical modeling and the visualizations they use to convey their results. It’s a great example of visualization applied to visualization research.

Week 1-2 Assignments

• Content Survey 1: Building Blocks (due Fri, Sep 25)
• Design Warmup 1: Ask Questions, Sketch Answers (due Fri, Sep 25)

Week 1-3 Sep 28-Oct 2 (Module 1)

For the last week of the module, you’ll focus on Design Exercise 1: Sketches and Questions (due Fri, Oct 02), so there is no reading. We’ll use class time to explore the building blocks more deeply and try some ICAs to help us understand them better. Design Exercise 1: Sketches and Questions (due Fri, Oct 02) will have you sketch ideas to solve some visualization problems.

Week 1-3 Assignments

• Anonymous Survey 1: Attitudes and Preferences (due Fri, Oct 02)
• Design Exercise 1: Sketches and Questions (due Fri, Oct 02)