Makeover Monday 2019 Week 5: Digital Economy and Society Index

Makeover Monday

The Digital Economy and Society Index (DESI) measures the digital performance and competitiveness of countries in the European Union. The original chart for this week’s makeover shows the 2018 DESI ranking of EU countries.


What works well?

  • Countries are ranked in order of total index score.
  • Bars allow easy comparison of total score.

What could be improved?

  • Stacked bars make comparing individual indicators across countries very difficult due to the lack of a common baseline.
  • Colours are overly vibrant. A more muted palette in the same tone would be less distracting.
  • Countries are not obvious based on two-letter country codes. The country names would be better as labels.

What Marc did:

  • Pivoted data to show each indicator in its own column in addition to the total score.
  • Created a table in Data Studio, allowing sort order to be changed by clicking on column headers.
  • Added reference line for the EU average of each metric.
View in Data Studio

What Ryan did:

  • Created slope graphs for total weighted score and for each indicator, comparing 2014 to 2018 ranking.
  • Used a highlight parameter to compare individual countries.
  • Added dimension descriptions from the source site to give context.
View in Tableau Public

Makeover Monday 2019 Week 4: Energy Use at 10 Downing Street

Makeover Monday

This week’s Makeover Monday looks at electricity usage at 10 Downing Street, the official residence of the Prime Minister of the United Kingdom.

What works well?

  • A time series chart is appropriate for displaying the trend in energy usage.
  • Peaks in usage are highlighted in lighter green.

What could be improved?

  • The graph does not provide a legend for the different shades of green. However, it is explained in the notes available on the source website.
  • It is difficult to see patterns in usage by hour of day.

What Ryan did:

  • Started off with a heat map by month and day of the week.
  • I also wanted show the trends by hour, so I created small multiple line charts also by month and day of the week.
  • Used transparent sheets to overlay the line charts on top of the heat map.
View in Tableau Public

What Marc did:

  • Created a calendar layout using week number and weekday in Tableau.
  • Charted hourly electricity usage as a heatmap; brighter spots indicate higher usage.
  • Used the Yellow-Green-Blue (YlGnBu) colour palette, available from Jacob Olufska’s Colour Palettes for Tableau viz.
  • Took some overall design inspiration from Justin Davis’ At Minimum viz.
View in Tableau Public

Makeover Monday 2019 Week 3: Minimum Wage Earners in America

Makeover Monday

This week, we tackle both Makeover Monday Week 3 and the January Storytelling with Data Challenge. For Makeover Monday, we visualize the percentage of U.S. workers earning minimum wage. For the SWDChallenge, we try using new data viz tools (Google Data Studio for Ryan, Microsoft Power BI for Marc).

The original chart for Makeover Monday shows the percentage of hourly paid workers in each state that earn minimum wage or less for 2017.

What works well?

  • The monochromatic, 4-colour scale is aesthetically-pleasing and easy to interpret.
  • Showing the data on a map makes geographic patterns stand out, in particular the belt of darker blue stretching from Louisiana to Virginia.

What could be improved?

  • Smaller states in the northeast are imperceptible on the map. Using a tile map would give all states equal prominence.
  • Data is shown for only the most recent year. It would be interesting to see how the percentage has changed over time.

What Marc did:

  • Created a tile map. Each chart shows the trend in the percentage of workers earning minimum wage or less for the respective state.
View in Power BI

In the spirit of trying something new for the #SWDChallenge, I decided to use Microsoft Power BI. It was easy to get started with Power BI Desktop, which has a built-in connector to data.world. After getting the data imported, it took only a few clicks to create my first chart.

The Power BI interface is intuitive, adopting the same ribbon menu as Microsoft Office products. The “visualizations” and “fields” configuration panes are also similar to the data and style panels in Google Data Studio.

I found the chart formatting options sufficiently flexible. It was easy to set the axis ranges and hide the axes. I appreciated the built-in title field for every chart. Arranging charts on the canvas was a bit frustrating; although there is a snap to grid feature, it doesn’t apply when multiple objects are selected.

Overall, I succeeded in created the viz that I envisioned, without any significant hurdles. However, I have only scratched the surface of what Power BI can do. Perhaps I will return to Power BI for some future vizzes.

What Ryan did:

  • Charted the overall percentage of workers earning above min. wage, below min. wage, and the total. I avoided doing state comparisons as there are different state level minimum wages.
  • I liked how easy it was to create smooth lines in Data Studio, but I could see that the trends were being misrepresented so I went with the regular lines.
View in Data Studio

I used Google Data Studio for this month’s #SWDChallenge. I’ve pretty much only used Tableau for data visualization thus far, so it was great to learn another tool.

Data Studio is easy to navigate and I liked how the interface looked. In particular, I liked the grid layout which helped with formatting and how easy it was to create smooth lines. I also enjoyed the ability to copy and paste text boxes which isn’t possible in Tableau. Some formatting that is easy in Tableau, such as labeling start/end of lines or hiding one axis, I wasn’t able to figure out in Data Studio.

Although this is a simple chart, I like the interactivity with the highlighted marks and tooltips. I look forward to seeing how Data Studio improves in the future!

Makeover Monday 2019 Week 2: Global Press Freedom

Makeover Monday

This week, we look at freedom of the press around the world. The map below categorizes each country as Free, Partly Free, or Not Free, as evaluated by Freedom House in their 2017 Freedom of the Press report.

What works well?

  • Colours are distinct and complementary. The green/yellow/purple palette coordinates well with Free/Partly Free/Not Free.
  • Using a map here is appropriate for communicating the geo-spatial relationships between countries and their level of press freedom. For example, it is immediately evident that most Not Free countries are in Africa and Asia, while the majority of North America and Europe have Free press.

What could be improved?

  • Instead of discrete categories, the press freedom score could better be shown as a continuous gradient. The underlying score is on a scale of 0-100; Not Free consists of scores from 61-100. There is a big difference between 61 and 100, yet this is impossible to see with the 3 categories shown above.
  • The map provides no indication in the change in press freedom over time. Is it getting better or worse?

What Ryan did:

  • Used the Global Peace Index for 2008-2016 (used for Makeover Monday 2016 Week 40) to identify a relationship with the Press Freedom scores. Some countries which had a Press Freedom score were not included in the Global Peace Index so these were excluded from the viz.
  • Created a scatter plot of the Global Peace Index vs Press Freedom scores for 2016.
  • Coloured the marks by region.
  • Sectioned the chart background according to the press freedom status using reference bands.
  • Used set actions and transparent sheets to show the progression for each country when a mark is clicked.

What Marc did:

  • Focused on the ten “Countries to Watch” as indicated in the original article.
  • Plotted the trend in the press freedom score for each country as small multiple time series charts.
  • Used scorecards to show the latest score for each country.
  • Borrowed the newspaper-like formatting from my earlier viz, The Changing Face of Major League Baseball.
View in Data Studio

Makeover Monday 2019 Week 1: NHL Attendance

Makeover Monday

The first Makeover Monday of 2019 offers us this chart of average attendance by team in the NHL.

What works well?

  • Teams are sorted in alphabetical order by city

What could be improved?

  • The y-axis for the bars should start at zero. Truncating the axis results in inaccurate comparisons.
  • A line chart should not be used across discrete categories. There is no trend here; the pattern of the line is determined by the ordering of the teams.
  • Showing the % change on a secondary axis can impart unintentional relationships. For example, the % change markers for Dallas and Phoenix appear above their respective attendance bars; this has no meaning. The position of the change marker relative to the bar is just the result of the axis scales.
  • 3D effects should be removed. They add nothing to the chart.

What Marc did:

  • Created small multiple charts by team.
  • Grouped the teams into divisions and conferences.
  • Charted average home attendance per game for each season since 2000.
  • Denoted Stanley Cup winning seasons, as these likely correlate with higher attendance.
  • Added annotations for Atlanta Thrashers relocation to Winnipeg and Vegas Golden Knights inaugural season.
Click to view in Data Studio

What Ryan did:

  • Calculated the percentage of arena capacity filled for home games (Credit to Justin Davis for his arena capacity calculation). I thought this was a better comparison as arena capacity varies among teams.
  • Created small multiple area charts for each team and grouped them by conference.
  • Coloured the charts by team colours.
  • Labelled the team names inside the area charts
  • Added a constant line at 100% capacity. I noticed that some attendance figures go over capacity which may be due to extra tickets sold or participation in the Winter Classic game.
Click to view in Tableau Public