Data Visualization

(Good, Bad & Ugly)

click for slides: bit.ly/uml_dvgbu

Outline

Hanging Rootograms?!

source: https://datavizproject.com/data-type/

What are we focussing on?

• Information visualization.


• Display data in a way that makes it easier for your audience to explore or understand your project.

We don't want 'chartjunk' (Tufte 1983)

1. Do you know your data

2. Who is your audience

Audience- Purpose

Exploratory: Population Estimate Dashboard

Explanatory: Assisted Suicide Infographic

Font- What about ugly?

• Font has personality, select combinations appropriate for your audience.


• Keep things simple: one decorative, one for body text.

Use a tool like FontPair help with font selection.

3. What are best practices in colour selection for classification

Colour- Continuous vs discrete data

Select color schemes appropriate to your data type (shade vs. hue).

Colour- What about ugly?

• Keep things neutral with 2-3 accent colours maximum.

Use a tool like Color Picker help with colour selection.

4. What are some unconscious perceptions we have around colour

Colour- Unconscious perceptions

1. Red is bad, green is good;


2. Light blue shapes on maps are water;


3. Light colours represent less, dark more.

Note:

Be engaged (and beware of 'universal truths').

Be engaged (1)

1. Data visualization theory is a colonial construct;


2. Colour has connotation outside of your own world view;

Be engaged (2)

1. "Maps have killed more people than guns ever have" Dayrit (2020)


2. Who are you focussing on? Who is being erased?

5. How can we take accessibility into consideration when we are selecting colour palettes

Colour- Accessibility:

source: (Manitoba Public Health, 2017)

6. What is cognitive load, and why does it matter in visualization

Cognitive load- Special Effects & Stacking

• Humans are awful at interpreting 3D graphics and clutter.


• 3-5 classes are ideal, 5-7 at most.

7. How important is consistency

Cognitive load- Consistency

Keep consistency between legends and graphics.

source: (Mandelbaum et al., 2012)

Cognitive load- Consistency

Keep consistency between axes, don't force correlation.

8. What can I do to improve the clarity of my graphics

Cognitive load- Clarity

Use explanatory text and colour to create emphasis.

Cognitive load- Clarity

Use small multiples.

Cognitive load- Clarity

Keep it simple (and don't mislead the user).

9. What technology considerations should be taken into account

Technology- Reproducibility

Standard file formats and programs.

Human component

Technology- Considerations

10. How can I get started

Resources:

Data Visualization LibGuide

Don't forget about training resources!

Key points:

Message - be selective

Audience - you're creating this for them

Data - requirements and structure

Be kind to your future self.

Questions