The world of visualization has a few different camps of thought, which can be roughly grouped as follows:
- Those who condone best practices at all cost.
- Those who are driven by paying clients.
- Those who just want to inform effectively.
To say that everyone in these three groups gets along perfectly would be a misleading statement, at best. These days, tension between them are almost a given. In hopes that a little bit of background may help us keep that tension at healthy levels, here’s the story of how it came to be, and why. Once upon a time, there was a fledgling field of study called Data Visualization. The field had actually started in the middle of the 18th century, but as with all fields of study, it was taking a long time to grow up and fly. Towards the end of the 20th century, though, major technological improvements gave it the wings it needed to really take off. Chief among them: computers finally got the graphical capabilities and capacity needed to automate the translation of numerical data into visuals. This brought immense power to the field and it finally started to get appreciation from other disciplines. Up until that point, visualizations had been difficult and time consuming to make. There weren’t too many of them created, and those that were drawn had to do double duty. They let their creators see into an otherwise opaque dataset, and they provided a good medium for presenting those findings to others. Because so few of them were created, the attitude towards them was one of close examination and long pondering, and the Internet’s culture of rapid viewing and sharing did not yet exist. The finished product provided instant gratification with a clear overview, while the lengthy creation process allowed for reflection and discovery. In short, data visualizations had provided a medium for both analysis and presentation. Now that computers could automate the process, data visualization started becoming more common. The ability to (relatively) quickly create new visualizations finally got it some recognition. Statistical programming software began to provide tools for creating charts from data. Spreadsheet software allowed people to see charts without ever programming. It wasn’t long before statisticians and journalists (both working in much more mature fields) picked up some of the results and ideas from data visualization. Perhaps this is where the perceived split between data visualization for storytelling and data visualization for analysis happened. Statisticians began using the ability to rapidly create charts as a part of their analysis process. It wasn’t long before Visual Analytics was formed, and software like ggplot2 and Tableau were created to speed up the analysis process even further. The increased data collection and processing capabilities that computers were bringing the world meant that journalists started incorporating data into their stories, too. They needed ways to show this data and call out specific patterns and points that the data contained. Graphic designers and artists were employed to turn raw digital output into more attractive visuals, and Data Journalism was born. This division of data visualization uses in the private sector started to have an impact on the blogging community and academic world. Critiques of work on each side were written, and papers were devoted and categorized to each subject. Tensions rose as each side (still too young to be confident in their own merits) analyzed and criticized the other. The tension only grew more palpable as infographics started gaining popularity on the internet. Marketers got their hands on data visualization and realized how powerful it can be for promotion, especially in the new sharing economy. Their work was similar to that of journalists, but rather than placing data visualizations into the journalist’s medium of articles, they placed them into advertising. Advertising lives right on the edge of truth, and this is a very dangerous place for a troubled field of study to be. So far, data visualization had built its reputation on the feelings of trust and respect that come from seeing data drawn out before your eyes. If too many data visualizations fell into the abyss of marketing lies, they could discredit the entire field, destroying that reputation.
Let’s take a moment to step back and look at the motivation of each of the parties so far.
- Statisticians primarily want to use visualization as a way to gain insight into data.
- Journalists are driven by informing the public and telling a compelling story.
- Marketers typically want to help their clients sell more of their products.
These motives tell us about the target audience of each group. Statisticians mostly create visualizations for themselves, while journalists and marketers make visualizations for others. There are differences between the audiences of marketers and journalists, though. The audience journalists have is reading out of interest and deep curiosity, and they probably came to the piece intentionally. The audience marketers have was fed the piece, and are viewing it because it happened to be flashy enough to catch their eye. The motives also tell us why each format has grown into what it has. Statisticians don’t need superfluous elements on the page, they just want to get the insight and move on. Journalists want a clear story told, but they also need to engage their viewers to make a strong impact and, to a degree, entertain. Marketers need POP! Their work depends on snagging the unsuspecting eye — and being flashy is the best way to do that. None of these disciplines can be faulted for using data visualization in their work. Data visualization is a powerful tool and it was bound to attract people from all kinds of disciplines. Each side does need to respect the other, while still remaining respectfully critical. Marketing especially, should be watched carefully: it lives in a dangerous neighborhood. But all three should maintain their integrity and respect the power of the tool they use. We should also recognize that there are no hard lines between the three fields. There is a continuum that each visualization sits on, and picking the precise boundaries between the three branches is difficult. Visualization is well beyond the point where it has to worry about falling to the ground. Tensions have cooled recently, as members of the data visualization community have begun to see the merits of each discipline. Data Visualization’s wings are strengthening, and the only question now is how high it can fly. Drew Skau is Visualization Architect at Visual.ly, and a PhD Computer Science Visualization student at UNCC with an undergraduate degree in Architecture. You can follow him on twitter @SeeingStructure