With the advent of new technologies in recent decades, historians have changed how they present their data. From simple bar graphs and pie charts to the eye-catching designs of Stream graphs, layers of information can now be dissected in one visualization. But, like their predecessors, these new methods of data visualization have their own problems which can scare users away for more traditional graphs.
Stream graphs were first developed by Lee Byron for a class assignment while a student at Carnegie Mellon University. The students were tasked with collecting and displaying a set of data in an interesting and novel way (Lee Byron & Martin Wattenberg, Stacked Graphs—Geometry & Aesthetics, 2). Byron collected the music listening trends of his classmates and created a flowing graph.
The thickness represents the number of songs from the artist played in a given week. The color saturation shows the overall number of times an artist is listened, and the hue represents the earliest date the artist was heard. The goal of this graph was to create something that was aesthetically pleasing and did not look mathematical or scientific. An unanticipated result of this was that the students were able to locate certain life events within the stream. “One pointed to the beginning and end of three separate relationships, and how his listening trends changed dramatically. Another noted the moment her dog had died, and the resulting impact on the next month of listening” (Byron, 2). The success of this visualization went viral and many imitators emerged. It can be hypothesized, according to Byron, that stacked graphs allow for large amounts of data to be shown to a large audience in a pleasing way.
The greatest success story of Stream graphs comes from The New York Times in which two-decades of movie revenue is shown. The strength of this type of visualization is apparent in the article. Online, users can use a scroll bar to travel through time and see each movie and its details. In print, there is only so much room to allow for data to be broken down. Each movie cannot be dissected and instead readers are forced to simply view the data with little interaction. The design can also pose a problem. With such a wavy design mixed with colors of varying hues, understanding data can take longer than other visualizations such as the line graph.
As shown with both Listening History and the Box Office Revenue graph, Stream graphs are used most effectively with changes over time. Changes over time in listening habits was the central theme of Listening History. The Box Office Revenue graph shows a similar change over time in regard to the popularity of movies over the life of their time in theaters.
Further reading: Stacked Graphs – Geometry & Aesthetics
Perhaps one of the most recognized data visualizations is a line graph. Similar to a scatter plot, line graphs require points of data to be placed on an (x,y) grid. But the two graphs differ in multiple ways. The first being that the data in a scatter plot does not necessarily need to be ordered. The benefits here are that outliers and trends can be easily identified.
In a line chart, the data is ordered typically by their x-axis value. The second, and most obvious distinction, is that the data in a line chart is connected by a straight line.
In the scatter plot, points of data are entered without an ordered system. A Line of Best Fit is usually entered as close to all points as possible. This is to show how the data would look if condensed to a single line. The line graph is most often used to show change over time connected by straight lines. These line segments allow for local changes—that is between a pair of points—to be easily shown. Because trends are shown, hypotheses for the future can be deduced by line graphs.
In this graph, seasons five and six of Game of Thrones is shown (as of June 6, 2016). The chart allows for the comparisons of both seasons’ episode by episode of total viewership. At the time of this graph, three episodes had yet to be aired which is represented by the dashed lines to predict future demand. With this, season 6 is more popular, by an average of 53% (Parrot Analytics).
Both a positive and negative of line graphs is their ability to display a lot of information in a single set.
Within this info graphic, the demand of Game of Thrones is clear during each season, highlighted in yellow. But between the seasons, there is an obvious decline in demand. However, there are peaks in the data that can be attributed to specific announcements and events involving future seasons. With each spike in demand, an announcement (such as future seasons) or event (Comic-Con) is clearly responsible. However, while it is important that these fluctuations can be visualized over the course of a year, the large amount of data makes it difficult to look at individual days. Without the added context of events and announcements, it would be hard for someone to generate cause and effect conclusions.