This article is part of an Essential Guide, our editor-selected collection of our best articles, videos and other content on this topic. Explore more in this guide:
2. - Data visualization tools evolve with big data: Read more in this section
- Achieving big data analytics success with data visualization tools
- Data visualization tools offer more capabilities -- and challenges
- Users feel sense of ownership when they can visualize big data
- Big data propels evolution of data visualization tools
- The importance of visualization tools for big data analytics
Explore other sections in this guide:
- 1. - Flash vs. function in dashboard design and development
- 3. - News and trends in dashboards and data visualization
- 4. - Dashboards and data visualization terms, defined
Keeping Up with the Quants: Your Guide to Understanding and Using Analyticsis analytics guru Tom Davenport's latest exploration of big data and analytics. Due out in June, the book offers a management primer on surviving what Davenport and co-author Jinho Kim call "the quantitative information age."
In part 1 of our New Books interview, Davenport talked about the importance of seeing data as a means of influencing business decisions. In this second part, he expands on the means to this end, discusses why visualization tools are critical but still quite under-developed -- and, oh yes, surmises that one day soon, we'll all need to become YouTube directors.
What is the fundamental problem with communicating analytics?
Tom Davenport: A lot of people are not that quantitatively oriented. And let's face it, analytics can be very complicated, both from a statistical and mathematical perspective and in converting that implication into action. It is not an easy thing to do. Our educational institutions have not helped us out too much in that regard, in that they don't really spend a lot of time teaching people how to describe or translate analytics for decision makers. Even on the visual side, my impression is there is not even a clear inventory of what are the possible ways to display analytics visually.
You give two interesting examples of how data visualization -- or lack thereof -- can matter greatly in disseminating important ideas: Florence Nightingale, who used pie charts to communicate her findings on infection and won great acclaim, and Gregor Mendel, the father of modern genetics, whose work went unrecognized for decades in part because of its obscure presentation.
Davenport: I thought some classical mentions might help people see this is a longstanding problem. I heard recently of another example involving Alfred Russel Wallace, who was a biologist of sorts at the same time that Charles Darwin was around. Wallace came up with very similar solutions of evolution and natural selection, and his book actually came up before Darwin's did but didn't get the attention it deserved, largely for reasons of PR. I have concluded that our scientists, statisticians and mathematicians all need PR people to work with them now.
SearchCIO New Books
Keeping Up with the Quants: Your Guide to Understanding and Using Analytics
Authors: Thomas H. Davenport and Jinho Kim
Publisher: Harvard Business Review Press
Publication date: June 11, 2013
Is the reason that data visualization is so important to communicating ideas because we are visual creatures, or is there something special about data in a visual form, as opposed to written words or auditory streams?
Davenport: I think visual presentation is powerful for a lot of people. We are all, I guess, except for the blind, visual creatures. But some people interact better with text and even numbers than [with] visual. I don't know -- and I am not sure anybody knows -- what the distribution of preferences is, but I know I am not much generally a visual person. I'd rather see a narrative-based description. I think in many cases we benefit from multiple ways of telling a story visually and in terms of narrative, so you want to have as many going on as possible in order to find one that engages the consumer.
I was wondering whether the fact that a picture can convey so many points of data at once -- parallel streams of information -- somehow makes it more effective for a lot of, if not all people.
Davenport: Again, I would say that it is all very contingent. As we know, there are good visuals and there are bad visuals, and we've all seen a lot of bad visuals in PowerPoint, for example. Edward Tufte has made a pretty good living demonstrating there are really bad approaches to visuals that often occur in PowerPoints. I think we are probably a lot farther along in the world of narrative -- we've been doing that for thousands of years -- than we are in visual displays of information. I think we are really just finding our way now.
I read an interview with Amanda Cox, who is a statistician by background and who does interactive graphics for The New York Times. She basically said we don't know very much about what we're doing. Design doesn't seem to provide a huge amount of insight to visual analytics. She says that she tends to find that people trained in computer science are the best at it, which I found somewhat surprising. So, I think we're in [the] really early days of knowing what the best approaches are and what kinds of skills are necessary to excel at the visual display of information.
Our educational institutions … don't really spend a lot of time teaching people how to describe or translate analytics for decision makers.
professor, Harvard Business School
What about some of the visualization tools out there -- are they up to the task?
Davenport: First of all, there is rapid change in these visual tools. You have companies like Tableau that are very popular now that were not popular at all five years ago. I think it is very confusing to CIOs and to people who have to make software selections, because you have some tools from SAS and IBM, which are traditionally good at statistical analysis but focus less on visual display, and then you have tools like Tableau and Spotfire that were much more oriented to visual display and less oriented to statistics. And now everybody is trying to augment those specialties: SAS has a new visual analytics approach, and Spotfire has a new statistical set of capabilities that they acquired. It is a just a very confusing landscape right now for people trying to make decisions.
It's hard to find instruction in this -- I don't know that there is an educational program anywhere in visual analytics. There's computer science and there's statistics and there's design, but it's the combination that we need now.
Final question: In all your research, what example of communicating analytics blew you away?
Davenport: I think probably the example of using video to illustrate analytics. This was a pretty straightforward financial reporting application, but the idea that they would augment it with a very creative video story is probably the way we're going to do things in the future.
If you look at how we are persuaded to do other things, to buy products, to amuse ourselves -- think of all the clever cat videos people watch on YouTube these days -- it can't be very long until we get into the approach of communicating the results of analyses through video. So, we've got to become, not so much reporters or game designers, but all become video directors if we're going to be good at this.
Let us know what you think about the story; email Linda Tucci, executive editor.