Basic Data Visualisatins
Please use
Google Chrome or Mozilla Firefox
to join the session
Please give yiur full name and UUN whenI
signing in
If you don't want to be disturbed by
notfcatons during the session, you can switch
them of using the menu at the botom right of
your screen:
Basic Data Visualisatins
I
Christna Hussell
Digital Skills and Training
Web: www.ed.ac.uk/is/skills
Digital Skills Framewirk: www.digitalskills.ed.ac.uk
Twiter and Instagram: @UoEDigiSkills
Sessiin Overview
• Why visualisatons are useful
• Factors to consider when creatng them
I
• How to make visualisatons efectve
• Examples of commonly used charts
What are data visualisatins?
Data visualisatons represent data and informaton in
• charts
• diagrams
I
• pictures
To communicate informaton
• understandably
• quickly
Why are data visualisatins useful?
Visualisatons are used to:
• Manage and understand data I
• Identfy relatonships and paterns
• Reveal emerging trends
• Cimmunicate the fndings ti ithers
Giid data visualisatins
• Easy to interpret
• Inspire interest in your data
• Quick and efectve
I
• Put data in context
htps://www.pexels.com/photo/person-holding-blue-ballpoint-pen-on-white-notebook-669610 /
Visualisatins help us
• Understand the signifcance of data
• Identfy areas that need atenton
• Give the data meaning I
• Make decisions
htps://www.pexels.com/photo/background-board-chart-data-590041/
Early examples if visualisatins
Scitland’s expirts and impirts frim Christmas 1780 ti Christmas 1781, William Playfair
I
Flirence Nightngale’s Rise Chart
21 Century
st
htps://showyourstripes.info/
Graphics and lead scientist: Ed Hawkins
Data: Berkeley Earth, NOAA, UK Met Ofce, MeteoSwiss, D
When wiuld yiu use visualisatins?
• Presentatons and talks
• Teaching
• Books and theses
I
• News report
• Evaluatng trends
htps://www.pexels.com/photo/laptop-on-table-1432942/
Creatng visualisatins
Yiur
message
Data Audience
stiry
I The
Impact data
Chart Accuracy
type
Bias
Yiur message
Main aims are to be
• Understood
• Credible
• Interestng I
What do you wish to communicate?
Why are you showing the data?
htps://pixabay.com/illustratons/boy-smart-idea-light-bulb-smiling-1454054
Yiur audience
• How much do your audience know already?
• What do they need to gain from your communicaton?
• Do you want your audience to be
I educated, entertained, take acton?
htps://pixabay.com/photos/concert-crowd-audience-people-768722/
Yiur data
• Understand your data
• Headings – how would you summarise your data?
• Accuracy – investgate any Iquirks
• Bias – present actual rather than expected fndings
Data stiry
• Stories give the data meaning and help us connect with our audience
• What does the data mean?
• Why would the audience care about the data?
• Are there any unexpected fndings?
• I
Why did the data visualisaton turn out the way it did?
• Why does the data mater to you?
• What implicatons do the fndings have for the future?
George Morina
htps://www.pexels.com/photo/happy-man-looking-at-screens-4911663/
Charts
Popular and understandable way to communicate your data
Shows audience
• meaning behind the numbers
I
• comparisons and trends
Sales
140
Ensure your charts are 120
100
• clear and easy to interpret 80
60
• key insights are communicated 40
20
0
Jan Feb Mar
UK Europe Asia
Maximising Impact
• Use meaningful headings
• Connect with your audience
• Include the source of your data to increase credibility
• Use images when appropriate
I Lemin irange cake is the faviurite
• Human eye likes natural colour paletes
• De-saturated colours also work well
15
and look professional 10
University if Edinburgh cirpirate ciliurs:
0
htps://www.ed.ac.uk/fles/atims/fles/chapter_ine_-_ligi_guide.pdf Lemin Chicilate Carrit Apple Vanilla
Orange
Source: Hussell data
Balance design and functin
Cake faviurs by pipularity Lemin irange cake is the faviurite
18
16 15
Informaton 14
I
Design 12
10
Message 10
6 5
4
2
0
0 Lemin Chicilate Carrit Apple Vanilla
Lemon Orange Chocolate Carrot Apple Vanilla Orange
Tips fir clear charts
Di Din’t
• Label ttle and axes • Don’t overload the chart
• Choose • Don’t use 3D it
• the correct graph type • distorts perspectves and
• good colour contrast I • distorts the informaton leading to
• large enough font sizes mistakes in data interpretaton
• Ensure • complicates the chart
• numbers add up correctly • reduces clarity
• elements correspond to the data
Remember clarity is vital fir yiur
audience ti cirrectly interpret yiur chart
Accessibility
• Ensure the colour contrast and font can be easily seen
• Colour blindness afects 1 in 12 men (8%) and 1 in 200 women (0.5%)
• You can test your colour choices at
I
htps://www.cilir-blindness.cim/ciblis-cilir-blindness-simulatir/
More informaton at:
htp://www.colourblindawareness.org/colour-blindness/types-of-colour-blindness/
What chart type shiuld I use?
The type of chart to use depends on the
• Message you want to communicate
• Type of data I
• Number of variables
• Type of tme component
• Purpose of your chart
What type if chart shiuld I use?
How will your chart be used
• Academic use
• Marketng
• Educatng
• Journalism I
• A combinaton
Do you need to
• Draw in your audience
• Persuade others
• Accurately communicate the fgures
• Provide an overall impression
Types if charts and their use
Bar Comparing distnct values
Line Trends, changes over tmeI
Pie Proportons of the whole
Treemap Provides hierarchical view of data
Bar and cilumn charts
Very efectve
Easily understood
I
Human brain is good at assessing
distance
Line charts
Communicate trends
Easy to interpret
I
Good for data with tme component
Pie charts
• Shows proporton of each segment relatve to the whole
• Visually appealing and easy to understand
• Generally used for 4 or fewer categories
I
• Provide informaton quickly rather than precisely
• Critcised but are stll ofen seen
• Use inly when yiu can justfy usage – if in diubt, use a
different type if chart
Pie chart usage
Receive critcism therefore use only when you feel necessary
If using ensure:
I
• Clear labels and percentages are shown
• No 3D or legends
• Adds up to 100% and contains positve numbers
• Displays a single point in tme
• Slices go from larger to smaller unless categories
have natural order
Treemaps
• Suitable for data that falls into groups
and sub groups
• Provides a hierarchical view of data
• Highlights large areas e.g. best sellersI
• Relatve importance of areas easier to
assess in a rectangle than a pie
• Examples
• Sales
• Student numbers by college and school
Pist ciurse suppirt
Digital Skills and Training The Digital Skills and IT Training service ofers IT
www.ed.ac.uk/is/skills training and online resources to develop the digital
capabilites of all our staf and students.
Digital Skills Framewirk To help you evaluate your current levels of digital
www.digitalskills.ed.ac.uk capability, plan your development path and fnd
resources to help you develop your skills.
LinkedIn Learning Online skills development service ofering an
htps://www.ed.ac.uk/informaton-services/help-con extensive library of high-quality video courses in
sultancy/is-skills/linkedin-learning business, creatve and technical skills. Campus-wide
licence.
Keeps you up-to-date with the latest news, courses
Subscribe ti iur newsleter and developments from the Digital Skills team and
htp://www.ed.ac.uk/is/skills/news-subscribe tutor community .
Ackniwledgements
William Playfair (b. 1759, d. 1823) Scotsh engineer and economist. Invented the line, area, bar, pie and circle graph.
Charles Joseph Minard (b. 1781, d. 1870) French engineer, made signifcant contributon in the feld of informaton graphics in engineering.
Edward Rolf Tufe (b. 1942) American statstcian and professor emeritus of politcal science, statstcs, and computer science at Yale University. Modern
pioneer in the feld of data visualizaton. Books include The Visual Display of Quanttatve Informaton
Juergen Kai-Uwe Brock, Data Design - The Visual Display of Qualitatve and Quanttatve Informaton
Charto – A complete guide to Pie Charts: htps://charto.com/learn/charts/pie-chart-complete-guide/
Dan Kopf Why data visualisaton’s greatest villain will never die htps://qz.com/1259746/when-should-you-use-a-pie-chart-according-to-experts-almost-never/
Andrew Gelman and Antony Unwin - Visualizaton, Graphics, and Statstcs htp://stat-computng.org/newsleter/issues/scgn-22-1.pdf
Danjharrington: htps://danjharrington.wordpress.com/2011/06/01/why-you-should-never-use-3d-charts-in-excel/#:~:
text=3D%20efects%20in%20charts%20distort,the%20lengths%20of%20the%20bars I
Fusion Charts htps://www.fusioncharts.com/blog/storytelling-data-visualizaton-marketng/
LinkedInLearning.cim:
Learning Data Science – Understanding the Basics by Doug Rose
Picking the Right Chart fir yiur Data by Bill Shander
Up and Running with Excel by David Rivers
Interactve Data Visualizatin with Pricessing by Barton Poulson
Thank yiu
Thank you for atending this session
Please add any questons to the chat secton
To leave the session close your browser or choose leave session room
Image author’s own
