3 ingredients of effective data visualization: audience, message, the right chart

The purpose of data visualization is to convey messages, not to awe audiences with spectacular visuals. A data journalist, or anyone who works with data visualization, needs to know who the target audience is, what information the audience wants, and how to choose the right chart.

Let’s say a writer is working on an article about U.S. economy for a U.S. media outlet, with a theme that United States was not dominating the world economy until early 20th century. A sub-topic is to compare U.S. economy with that of major countries in 21st century, and the writer wants to include some (imaginary) historical data on GDP of major countries in 1870:

gdp_numbers

If the writer just wants to “show numbers in pictures,” it is easy and popular to create a pie chart by loading the raw data to a visualization tool such as Excel or, in this case, infogr.am.

gdp_pie

A pie chart obviously looks better than a table, and it also conveys the message that U.S. GDP was but a small part of the global economy in 1870; in fact, this is what a pie chart does best – showing part-to-whole relationship.

However, the article itself has been elaborating at length that U.S. economy was a small one in 19th century, do we really need a pie chart to “repeat” that message? A data visualization needs to clarify or enhance the subject by providing additional information or insights, and a pie chart is not the best choice for that purpose.

If a U.S. reader is interested in knowing how U.S. economy performs in 1870, there are several things the reader wants to find out quickly, such as (a) whether U.S. GDP is larger or smaller than a specific country of interest, (b) how U.S. compares to rivals such as England, Japan, Russia, etc. (c) actual GDP numbers of some countries of interest.

Obviously, this pie chart is not able to provide such information. Looking at the pie chart alone, a reader has difficulty in telling if U.S. GDP is larger than, say, that of France, or if U.S. ranks higher than some comparable countries, and there is no way to find out actual GDP numbers.

So, if we know the target audience is U.S. readers, the information they want is how U.S. GDP compares to other countries, then we need to choose a chart that best communicates such information.

A customized bar chart, such as the one below, does a better job in (quickly) communicating to U.S. readers what they want to know:

  • U.S. has a GDP of US$98 million, which puts U.S. in the middle range of the major economies at the time.
  • U.S. is better than Germany and Russia, but is behind England and France.
  • U.S. GDP is about half that of China, but six times that of Brazil.

gdp_bar

FYI: To create this bar chart, I first made two changes in Excel: (a) added GDP numbers to country names and (b) sorted numbers in ascending order, then loaded the data to infogr.am, created a bar chart, and set colors of other country bars to gray.

About Mu Lin

Dr. Mu Lin is a digital journalism professional and educator in New Jersey, United States. Dr. Lin manages an online marketing company. He also manages MulinBlog Online J-School (www.mulinblog.com/mooc), a free online journalism training program, which offers courses such as Audio Slideshow Storytelling; Introduction to Social Media Marketing; Writing for the Web; Google Mapping for Communicators; Introduction to Data Visualization; Introduction to Web Metrics and Google Analytics.
This entry was posted in Data journalism. Bookmark the permalink.

One Response to 3 ingredients of effective data visualization: audience, message, the right chart

  1. Pingback: 3 ingredients of data journalism: audience, mes...

Leave a Reply