Episode 04: How to Find Answers in Survey Results - Featured Viz by Shirley Wu
Welcome to episode 4 of Data Viz Today. How can you pull answers from and visualization survey results? I share Shirley Wu's data viz where she visualized the results of a survey of 981 data viz creators' frustrations. Her analysis and visualization provide insights on how to find answers in survey results.
Welcome! I'm Alli Torban.
00:30 - I’ll be creating my own project to share each week that’s inspired by the featured viz. Try it with me! #InspiredByDVT
00:45 - Because of this, I’ll be switching to releasing episodes once a week - every Tuesday.
01:00 - Featured viz by Shirley Wu “655 Frustrations Doing Data Viz”.
01:15 - Inspired by a Survey of data viz creators by Elijah Meeks to see why people were leaving the data viz field.
01:45 - How can we find answers in survey results?
02:00 - Shirley Wu is a freelance software engineer and data viz consultant based in San Francisco.
02:15 - Shirley wanted to know what data viz creators’ frustrations are so the community can help keep them in the field.
02:25 - Download the data here.
02:30 - She separated the respondents into two groups: those who listed a frustration and those who didn’t.
03:10 - Shirley’s creative process is to sketch, code, sketch, code, repeat.
03:25 - Finally decided on a bee swarm plot and RJ Andrews suggested that she represent the respondents with frustrations dropping down the X-axis as if they’re crying.
03:50 - Whenever you can tie the data back to the human story, it becomes more powerful.
04:36 - Bee swarm plot tutorial in R by Nathan Yau.
05:15 - Shirley created 12 of these plots and compared the results to find those who are that the most dissatisfied data visualization practitioners are ones who are:
On an embedded team,
Weren’t initially hired to work on data visualization,
Work on data visualization as only one of several tasks, and
Have a subordinate relationship with the stakeholder of their visualization
05:50 - Community can help by educating people on how data viz can help organizations and keep creating tutorials.
06:10 - Shirley’s advice when you hit the inevitable snags - create a viz about a dataset or question that you’re really passionate about. Being curious to see the final viz will keep you going through the frustrating times. Also try to take breaks and come back with fresh eyes.
06:55 - What if you have to create a viz for something you’re not passionate or don’t have time for breaks? A study published in Social Psychology and Personality Science found people don’t suffer decision-fatigue as much when they’re making decisions for someone else because it’s more enjoyable to choose for another person. Another study by the same author revealed that people are also more creative when making decisions for other people rather than themselves. So next time you feel creatively stuck, think about someone whose work you admire - maybe someone with a strong point of view or style - and ask yourself - what would he or she do?
07:50 - My 3 main takeaways:
If you’re trying to visualize survey results, it can be helpful to quantify the qualitative answers by grouping them in some way.
Try to visualize the distribution of the survey answers to two questions, and then add filters so you can see patterns in how people who are answering question 1 then answering question 2.
Try to incorporate a human/emotional element to your viz. It can connect people to your visualization and help them remember it.
08:30 - My inspired viz:
09:50 - Alluvial diagram on rawgraphs.io.
10:50 - My human element was to color the flows that hints at skin tones so it looked like arms reaching out for help.
12:00 - Thanks to Shirley!
12:20 - She’s excited to start using WebWorker to improve performance in the browser.
12:37 - Next featured viz is by Guney Soykan called "Face of a Nation".
- 12:50 - Subscribe to my weekly newsletter, which compiles the top tips from the week into a short email.
- 12:58 - Nominate a data viz to be featured on the show.
- 13:02 - Contact me on social media @dataviztoday!
New episodes are released every Tuesday at 6 AM EST.