Effective visual communication

class: center, middle, title-slide

.title[
# Effective visual communication
]
.author[
### Claus O. Wilke
]
.date[
### last updated: 2023-02-17
]

---

## Why Data Visualization?

<br>

.center[
<img src = "introduction_files/Data_science_venn.svg", width = 60%></img>
]

---

## Why Data Visualization?

<br>

.center[
<img src = "introduction_files/Data_visualization_venn.svg", width = 60%></img>
]

---

## Why me?

.pull-left.small-font[

.center[
https://clauswilke.com/dataviz
]

]

---

## How did I get here?

---

## Three key principles of effective visual communication

1. Tell a story

2. Prepare figures that make a point

3. Make your figures accessible

---
class: center middle

## 1. Tell a story

---

## What is a story?

A story is a collection of observations, facts, or events presented in a specific order such that they create an .highlight[emotional reaction]

---

## Every story has an arc

.center[
Opening
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Challenge
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Action
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Resolution
]

---

## Every story has an arc

.center[
Opening
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

.highlight[Challenge]
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Action
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

.highlight[Resolution]
]

.absolute-bottom-right[
Challenge and resolution are<br>the two most important parts
]

---

## Mapping my story onto this format

**Opening:** We need to make many data visualizations in my lab

**Challenge:** I end up telling my students the same things over and over

**Action:** I go and write a book about data visualization

**Resolution:** The book is written. But now I teach the material, so I still keep saying the same things over and over

---

## Other story structures

.center.move-up-1em[
Action
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Background
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Development
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Climax
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Ending
]

.absolute-bottom-right[
(Better for action movies<br>than for scientific reports)
]

---

## Mapping my story onto this format

**Action:** In May 2017, I embark on a major project: Write a book on data visualization

**Background:** I do this because we need to make many data visualizations in my lab, and I end up telling my students the same things over and over

**Development:** Writing the book takes me almost two years; along the way, I add many features to ggplot2 and become a member of the ggplot2 team

**Climax:** The book is released in April 2019

**Ending:** Now I'm a dataviz teacher

---

## Other story structures

.center[
Lead
]

.center[
<img src = "effective-communication_files/arrow.svg", width = 3%, style = "margin-top: -0.5em; margin-bottom: -1em"></img>

Development
]

.absolute-bottom-right[
(Commonly used in newspaper articles)
]

---

## Mapping my story onto this format

**Lead:** Because I see the need for more education in data visualization, I have written an entire book about the topic

**Development:** I had previously written an R package to improve figure design, but it wasn't sufficient: Good judgement cannot be automated

So I wrote a book; the book is entirely about concepts, not about coding, and it is meant as a resource for anybody doing data visualizations, regardless of their preferred visualization software

---
class: middle center

## 2. Prepare figures that make a point

---

## How to use figures to tell a story

- Every figure needs a clear purpose within the story arc

- A complete arc will usually require at least two figures

---

## Example: Preprints in biology

.center[
![](effective-communication_files/figure-html/preprints-part1-1.svg)
]

.absolute-bottom-right[
Opening
]

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Example: Preprints in biology

.center[
![](effective-communication_files/figure-html/preprints-part1-1.svg)
]

<div style = "position: absolute; top: 50%; left: 57%; width: 3.8in; background: white;">
What happened in 2014 to curtail the growth?
</div>

.absolute-bottom-right[
Challenge
]

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Example: Preprints in biology

.center[
![](effective-communication_files/figure-html/preprints-part2-1.svg)
]

.absolute-bottom-right[
Resolution
]

<div style = "position: absolute; top: 65%; left: 60%; width: 3.8in; background: white;">
bioRxiv opened in 2014
</div>

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Don't overload your figures with irrelevant detail

.center[
![](effective-communication_files/figure-html/arrival-delay-vs-distance-1.svg)
]

Flight delays by airline, connection, and distance flown

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Don't overload your figures with irrelevant detail

.center[
![](effective-communication_files/figure-html/mean-arrival-delay-nyc-1.svg)
]

<div style = "position: absolute; top: 24%; left: 45%; backgro
und: white;">
American and Delta<br>have the shortest delays
</div>

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Make your figures memorable

---

background-image: url("https://clauswilke.com/dataviz/telling_a_story_files/figure-html/petownership-bar-1.png")
background-position: left 50% top 60%
background-size: 65%

## Make your figures memorable

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

background-image: url("https://clauswilke.com/dataviz/telling_a_story_files/figure-html/petownership-isotype-1.png")
background-position: left 50% top 60%
background-size: 65%

## Make your figures memorable

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---
class: middle center

## 3. Make your figures accessible

---

## These design elements all affect accessibility

- Font/symbol size

- Font choice

- Color choice

- Contrast

---

## These design elements all affect accessibility

- .highlight[Font/symbol size]

- Font choice

- .highlight[Color choice]

- Contrast

---

## Don't use tiny fonts!

.center[
<img src = "https://clauswilke.com/dataviz/small_axis_labels_files/figure-html/Aus-athletes-small-1.png", width = 60%></img>
]

.tiny[
Both text and symbols are way too small
]

---

## Don't use tiny fonts!

.center[
<img src = "https://clauswilke.com/dataviz/small_axis_labels_files/figure-html/Aus-athletes-good-1.png", width = 60%></img>
]

.tiny[
Text and symbols are reasonable but could be bigger
]

---

## Don't use tiny fonts!

.center[
<img src = "https://clauswilke.com/dataviz/small_axis_labels_files/figure-html/Aus-athletes-big-ugly-1.png", width = 60%></img>
]

.tiny[
Symbols too small compared to text
]

---

## Don't use tiny fonts!

.center[
<img src = "https://clauswilke.com/dataviz/small_axis_labels_files/figure-html/Aus-athletes-big-good-1.png", width = 60%></img>
]

.tiny[
Make fonts bigger than you feel comfortable
]

---

## Don't use tiny fonts!

.center[
<img src = "https://clauswilke.com/dataviz/small_axis_labels_files/figure-html/Aus-athletes-big-good-1.png", width = 40.8%></img>
]

.tiny[
Make fonts bigger than you feel comfortable
]

---
class: middle

## Rule of thumb:<br>Keep ratio of largest and smallest font size < 2.

---

## Design for color-vision deficiency

5%–8% of men are color blind!

Red-green color-vision deficiency is the most common

---

## Design for color-vision deficiency

5%–8% of men are color blind!

Blue-green color-vision deficiency is rare but does occur

---

## Design for color-vision deficiency

Choose colors that can be distinguished by people with CVD

---

## Consider using the Okabe-Ito palette

.center[
<img src = "https://clauswilke.com/dataviz/pitfalls_of_color_use_files/figure-html/palette-Okabe-Ito-1.png", width = 100%></img>
]

.tiny-font[
Name           | Hex code &nbsp;&nbsp; | R, G, B (0-255)
:----------    | :-------  | :--------
orange         | #E69F00   | 230, 159, 0
sky blue	     | #56B4E9   | 86, 180, 233
bluish green   | #009E73   | 0, 158, 115
yellow	       | #F0E442   | 240, 228, 66
blue	         | #0072B2   | 0, 114, 178
vermilion	     | #D55E00   | 213, 94, 0
reddish purple | #CC79A7   | 204, 121, 167
black	         | #000000   | 0, 0, 0
]

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## CVD is worse for thin lines and tiny dots

.move-up-1em.center[
<img src = "https://clauswilke.com/dataviz/pitfalls_of_color_use_files/figure-html/colors-thin-lines-1.png", width = 65%></img>
]

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Color encoding alone is not always sufficient

.move-up-1em.center[
<img src = "https://clauswilke.com/dataviz/redundant_coding_files/figure-html/iris-scatter-one-shape-1.png", width = 75%></img>
]

Iris virginica and Iris versicolor are difficult to tell apart

---

## Rearranging colors and using different shapes helps

.move-up-1em.center[
<img src = "https://clauswilke.com/dataviz/redundant_coding_files/figure-html/iris-scatter-three-shapes-1.png", width = 75%></img>
]

---

## When in doubt, run CVD simulations

<div class = "move-up-1em" style = "width: 50%; box-shadow: 3px 5px 3px 1px #00000080; padding: 8px;">
.small-font[Original]<br>
<img src = "effective-communication_files/iris-scatter-one-shape-edited.png", width = 100%></img>
</div>

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## When in doubt, run CVD simulations

<div class = "move-up-1em" style = "width: 50%; box-shadow: 3px 5px 3px 1px #00000080; padding: 8px;">
.small-font[Original]<br>
<img src = "https://clauswilke.com/dataviz/redundant_coding_files/figure-html/iris-scatter-one-shape-1.png", width = 100%></img>
</div>

???

Figures from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## When in doubt, run CVD simulations

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## When in doubt, run CVD simulations

???

Figures from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Pay attention to legend order

.center[
![](effective-communication_files/figure-html/tech-stocks-bad-legend-1.svg)
]

Legend order **does not** match visual order of the lines

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Pay attention to legend order

.center[
![](effective-communication_files/figure-html/tech-stocks-good-legend-1.svg)
]

Legend order **does** match visual order of the lines

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Even better: Don't use a legend at all

.center[
![](effective-communication_files/figure-html/tech-stocks-good-no-legend-1.svg)
]

This technique is called "direct labeling"

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Three key principles of effective visual communication

1. Tell a story

2. Prepare figures that make a point

3. Make your figures accessible

---

## Further reading

- Fundamentals of Data Visualization: [Chapter 29: Telling a story and making a point](https://clauswilke.com/dataviz/telling-a-story.html)

- Joshua Schimel (2011) Writing Science: How to Write Papers That Get Cited and Proposals That Get Funded.

- Zeileis et al. (2020) [colorspace: A Toolbox for Manipulating and Assessing Colors and Palettes.](https://www.jstatsoft.org/article/view/v096i01) Journal of Statistical Software 96:1–49.

- Fundamentals of Data Visualization: [Chapter 19: Common pitfalls of color use](https://clauswilke.com/dataviz/color-pitfalls.html)

- Fundamentals of Data Visualization: [Chapter 20: Redundant coding](https://clauswilke.com/dataviz/redundant-coding.html)