List the north and south cardinal points of the color wheel
Red: 0°, Cyan 180°.
[yellow 60°, green 120°, blue 240°, magenta 300° ]
Describe the indicators of intensity in color
- Saturation (chroma), how much does the colour diverge from white.
- Lightness (brightness), how luminous you perceive the colour to be.
Explain how we perceive intensity of colour.
We see the difference between objects, i.e. we perceive relative intensities well whereas we are not designed to measure their absolute values.
True or false: full-colour saturation should be used when possible, to make it clear what you are representing
False, too much saturation will cause clutter and cognitive overload.
Is colour a preattentive attribute?
No, it’s a channel or visual attribute. Color hue and intensity are preattentive attributes
Which of these statements on colour is correct?
- it’s effective in drawing attention.
- Using multiple colours allows the viewer to focus better on the visualization.
- Using a grey background and colour as a highlighter works well, in particular, if you choose red.
- A diverging colour scale should be used with categorical data.
- Colour is also a good way to decorate and make your visualization eye-catching.
- Being consistent with the use of colours is important.
The correct answers are 1 and 6.
List the common ways to use colour in data visualization and what to use them for.
- Sequential.
Colour is ordered from low to high hue, (white-blue, light-blue … dark-blue, navy), best with quantitative data. - Diverging:
Two sequential colours with a neutral midpoint, best to show how divergent the values are from the baseline on a positive/negative scale. - Categorical:
Contrasting colours for individual comparison, best for categorical or nominal (clusters) data. - Highlight:
Highlight some detail, best used when the rest of the background is neutral, for instance, grey. - Alert:
Get the reader’s attention and show that something is wrong, use fight or flight response colours, like red or yellow for western culture.
True or false: A sequential color scheme is always produced with one colour hue.
False, sequential palette can be single or multiple hue.
When is the rainbow scale best used?
- with numerical data like temperature
- as a highlighter in a table
- for categorical data
- never
Never.
- use 1 or 2 hues with a sequential palette for quantitative values
- highlighting a bunch of elements with different colours will create confusion instead of focusing attention
- using more than 6-7 colours for categorical data creates cognitive overload.
Which of these is a colour vision deficiency?
- Protanopia, red-blind
- Deuteranopia, green-blind
- Protanomaly, red-weak
- Deuteranomly, green-weak
All of them of course, did you have any doubts?
True or false. Standard green yellow red traffic lights take into account CVD
Wrong, unfortunately red and green are two of the most common CVD’s.
What can you do to make your visualization colour-blind friendly?
- Use a colour blind palette.
- Do not abuse colours for encoding.
- Check your final product with Sim Daltonism, and make sure the message is not lost if colours are not perceived the same way.
