Data visualisation

How good is your data?

• Make sure your data is clean.
• Check for duplication, missing data or information.
• Remember - Garbage In, Garbage Out.
• Is your visualisation open for misinterpretation due to bad or incorrect data?

Are you using the right kind of visualisation for your data?

• Consider what type of visualisation is best to convey your message.
• Think about your audience - will they understand the type you choose?
• What are you trying to show? e.g. relationship, comparison.

Static vs interactive?

• Do you want your data to be explored?
• How are you using your visualisation? e.g. presentation, website, journal article.

Be aware of clutter

• Is all the included information adding value? If not, remove it.

Is there too much information?

• Don't try to include everything in one visualisation. Consider using multiple ones.
• Be clear about what you are trying to say. 
• Focus on a specific point or option. This can help increase understanding².

How have you used colour?

• Limit number of colours used. Are they causing confusion instead of clarity?
• Don't choose a colour just because of aesthetic reasons. Could the colour have an unintended meaning e.g. due to cultural reasons?
• It should have a purpose e.g. highlight something, show related data.
• Don't use similar shades when trying to show contrast.
• If using a fill colour, use a dark shade.

Are the scales used inconsistent?

• Ensure the scale of multiple visualisations are consistent unless you want to indicate importance or emphasis.

Who is your audience?

• Don’t be too general or try to reach too many people.
• How are you perceived by your audience? Are you seen as a credible source for this information?
• Use different visualisations for different audiences.
• Consider what tone eg. serious, celebratory would be appropriate.

Labels, headings and text

• Are your labels providing clarity or confusion?
• Do you need to add labels or headings to help communicate your data?

Use of contrast

• Lack of contrast can make data less easy to understand.
• To highlight something, make it very different e.g. shading.

Be careful with alignment when using text

• Make use of clean lines.
• Try not to use too many diagonal elements.

Test it out

• Get feedback e.g. audience sample, colleagues, peers. 
• Fresh eyes can provide different perspective.
• Is it communicating your data the way you want it to?

Sources ^{3, 4, 5}

Colour blindness (or colour vision deficiency, or CVD) affects about "8 percent of men and 0.5 percent of women"⁶. The most common type is "red/green colour blindness". It makes it hard to decipher colour coded information, such as used in data visualisations.

Tools

Adobe Photoshop or Illustrator

• Can use Proof Setup tool to simulate how someone who is colour blind would see a visualisation.
• To access, go to View > Proof Setup > Color Blindness