data visualisation: art or science?
72hrs of you tube video
571 new websites
100m new emails
277,000 tweets
.. created every minute
Channel growth
Data vs Visualisation
Where do you start?
Data challenge: quality
Data challenge: understanding
Data challenge: outliers
Thinking about visualisation
Question. Question. Question.
› Male gender bias in graduate admissions.
› At department level: most departments had a small but statistically significant bias in favor of women
› Situation:∙ Women were applying to competitive departments
with low rates of admission ∙ Men tended to apply to less-competitive departments
with high rates of admission
Beware aggregate data
Source: Science, Bickel et al (1975)More info: https://www.boundless.com/statistics/statistics-in-practice/observational-studies/sex-bias-in-graduate-admissions/
4,321 applicants35% admitted
8,442 applicants44% admited
Beware the y-axis bias
March April May June July3.1363.1383.14
3.1423.1443.1463.1483.15
3.1523.154
March April May June July0
0.5
1
1.5
2
2.5
3
3.5
Going beyond the bar chart
Source: Nate Agrin and Nick Rabinowitz
Going beyond the bar chart
Timelines and geo
Source: http://hint.fm/wind/
Visualisation OverloadVisualisation overload
Remember the different channelsA step too far…
What we speak aboutbecomes the housewe live in- Hafiz
› Visualisation is both art and science.› The visualisation should inform not just be pretty› Lots of potential insight – start small with the most
important first› Don’t forget to understand and clean the data › Look at your outliers for potential new opportunities› Understand what you are trying to takeaway from the data
∙ Let that guide your choice of visualisation› Shy away from dashboard overload!› Don’t forget more often than not in marketing, your data
involves people
Key takeaways
Thank you
Top Related