Weekly Thoughts: Chart or Table?
Here is something that caught our eye this week:
Chart or Table?
Recently, we have been working on improving our reporting both at the individual company level as well as across the portfolio. As with any reporting initiative, the key objective of our work is to facilitate a repeatable aggregation of an increasing amount of raw data, the analytics of which allows management to obtain key insights into the health of the operations, thereby enabling informed business decisions. As one would expect, this initiative involves focused discussions on what data should be presented, in what format, and at what frequency. While we all agree on the overall objective, how to actually visualize the data is often an area of strong opinion and extreme debate. As such, we were interested this week to learn more about the evolution of data visualization and the benefits, and limitations, of common practices.
Interestingly, when thinking about ways in which to display data, the chart is a relatively new innovation, particularly when compared with the table, which dates back to the 2nd century. Furthermore, the bar, line, and pie charts were all invented in the 18th century by one person — Scotsman and political economist William Playfair — likely in response to a tremendous increase in the popularity of government-led economic and demographic data collection at the time. Initially considered childish by serious academics, it wasn’t until the 20th century that practitioners in government, business, and science regularly used graphical representations of data to summarize their work. As noted by content marketing firm Priceonomics, it wasn’t until 1933 that The New York Times regularly published a daily stock market line chart, now one of the most ubiquitous examples of graphical data visualization, as before then “there were lots of statistics in the paper. Journalists wrote about trends in the bond markets, activity in the commodity markets, and changing birth rates. Numbers were a big part of the media; they just weren’t visualized.”
As technology evolved to enable more efficient analysis of data sets (Lotus 1-2-3, Microsoft Excel, PowerPoint, personal computers, etc.), the use of graphical visualization has become commonplace for the presentation of all types of data as anybody who has attended a business meeting or read a memo in the past decade can attest. If you are wondering how commonplace, check out the Priceonomics chart below (see what we did there?).
There is no doubt that graphical visualization is an incredibly effective way to process and analyze patterns in large amounts of raw data. That said, there is some concern among data scientists that the use of charts is out of control. For instance, Andrew Gelman, director of the Applied Statistics Center at Columbia University, strongly favors tables over charts, arguing the use of graphs in research papers often leads to the implication of relationships that are not actually statistically significant, even when the underlying research is sound. Even outside of academia, there is concern that the use of graphs can create an impression of analytical rigor when in fact the implications may be misleading or the underlying data may be of poor quality. Gelman argues:
“Graphs are gimmicks, substituting fancy displays for careful analysis and rigorous reasoning. It’s basically a tradeoff: the snazzier your display, the more you can get away with a crappy underlying analysis. Conversely, a good analysis doesn’t need a fancy graph to sell itself. The best quantitative research has an underlying clarity and a substantive importance whose results are best presented in a sober, serious tabular display.”
At Chenmark, we are not abolishing all of our snazzy charts in favor of sober data tables. We continue to believe that graphical visualization of data can be an incredibly effective way to efficiently convey information. That said, our reading this week is a good reminder that during our current reporting initiative, we must not get too carried away with overly complex visual representations and risk creating the potential for misleading takeaways or insufficient analytical rigor. In sum, there’s a time and a place for sobriety, just as there’s also a time and a place for snazziness.
Have a great week,
Your Chenmark Capital Team