On Interactive Data Visualization, Illustrated

This past week the latest edition of New Directions for Evaluation started hitting mailboxes. The issue covers Data Visualization in Evaluation (it’s part 1 of 2). In it you’ll find an article written by yours truly on recent developments in quantitative visualization.

I tried to cover a few things, but the topic that has captivated me the most over the last few years has been interactive visualization. And I mean captivated. I’ve been pushing myself on the technical end so much that I’m technically a Systems Analyst now.

The natural constraints of a paper journal didn’t give me an opportunity to dive into the practical side of the subject, but that’s why it’s nice to have a blog. This post is the first in a series I have planned on data visualization. Call it an introduction to interactivity, hope you like it.

Social Media Makeover

Before we get into interactivity I wanted to let you in on a new idea I’d like to try out. I want to do a few social media makeovers, taking published data sets and PDF reports and reinterpreting them for the social media world. Depending on the set or report, you can expect cartoons, infographics, and interactive visualizations.

I need partners!

If you have a report or data set and would like a little free web dissemination support, here’s your chance. Send me an email: freshspectrum @ gmail .com. This isn’t a standing offer so act quickly.

My two requirements; the report must contain good evidence (qualitative, quantitative, whateverelseitative) and most importantly, it must be something you think is really important to share with the world. I’m not looking to decorate junk.

The numbers I care about

What we call a data set is just a collection of individual bits of data. We tend to analyze these bits in aggregate. For all the complaints about 200 page reports, they’re far easier to understand than 2,000 page reports.

But what we should remember, is that often those little bits can have far more meaning than the aggregate when presented to the right audience. Interactivity can allow us to share those bits efficiently and effectively.

Could you email me the numbers I actually care about

Visual Information Seeking Mantra

To call interactive data visualization a “recent development” is a little strange. We can find literature on the subject in the computer sciences going back decades. Just check out some of the academic work by Ben Shneiderman.

Even though it’s easy to find good examples today, interactive visualization is still a novelty in most fields. I think there are a few reasons for this…

Ben Shneiderman Visual Information Seeking Mantra

Paper is the default

The culmination of most evaluation and research projects is a paper report. Even though results are disseminated using the web, often that’s through identical digital versions of the paper report, downloadable in PDF form.
Paper is not interactive

Technical Manpower

Ben Fry, founder of Fathom Design and the creator of some sleek interactive visualizations, said it tongue-in-cheek in 2011 when discussing one of his major contributions.

Processing seeks to ruin the careers of talented designers by tempting them away from their usual tools and into the world of programming and computation.  Similarly, the project is designed to turn engineers and computer scientists to less gainful employment as artists and designers.

With all the demand for good computer scientists and engineers, is a career in data visualization a worthy endeavor?

Interactive visualization requires a mix of expertise in data analysis, graphic design, and programming (at least it has in the past). It’s tough to find individuals with competence across all three. Not surprisingly, the well known design shops tend to work in teams.

Ben Fry turning engineers and computer scientists to less gainful employment as artists and designers

Eyeo 2013 – Ben Fry from Eyeo Festival on Vimeo.

The Web’s Adolescence

We’ve come a long way in these past 18 years since the web went all commercial. But we have so much further to go.

Unlike news media, the ivory tower model has not been turned on its head. The publishing platforms of old are as popular and profitable as ever. Publish or perish has yet to extend to social media.

Social media changes the way we can disseminate evidence, but until web publishing becomes more than a hobby for a relative few academics, we will not see its full potential.

Adolescent web

Journalism’s influence

If there is a field leading the way in showing us what’s possible with interactivity, it’s journalism. Tableau, recognizing its appeal, even went as far as offering free versions of 2,000 dollar software to investigative journalists.
Journalists are the masters of speaking to large diverse audiences. Interactive visualization gives them another tool.
Amanda Cox, three things you can do with a data visualization

Personalizing the audience

When we write reports we tend to write for “states” or “counties” or “communities” as if they were a single entity. Anyone who has worked with “states” or “counties” or “communities” knows that they are certainly not a single entity.

Interactivity offers a pathway to reach many specific audiences.We live across the country do you think we're the same audience?

A different perspective

I’ve spent a good bit of my professional life staring at data sets, asking questions and running analyses. Playing with data is the best way to understand data. Of course, not everyone is comfortable staring at numbers in a table and asking questions with code.

Setting up your data in an interactive visual format can give your audience new perspective and an opportunity to answer their own questions.

Eric Rodenbeck If you ask questions about a data set what can you find?

Eric Rodenbeck speaks at TEDx Silicon Valley 2011 from TEDx Silicon Valley on Vimeo.

Practical and Affordable

I believe that interactive data visualization can be both practical and affordable. I hope to prove that to you (eventually).

Earlier on I asked for partners.  If you would like to work with me, shoot me an email freshspectrum @ gmail .com.

As always, looking forward to your comments below! And stay tuned for more data visualization posts in the coming weeks.

 

Comments

    • Santiago Ortiz says

      I apply many of the concepts featured in this article in my work: http://moebio.com The relations between overview, zoom and details, is something that I specially explore in this project: http://moebio.com/newk/twitter/ in which by clicking a node it’s possible to see its local sub-network (zoom) without loosing the context (overview); by dragging from one user to another you get the specific information (details) about the two persons (conversations if existing, intermediate nodes if not)… there are more than 500 000 possible pairs to select!

      Excellent post, thanks.

  1. Jeff Wasbes says

    Another winning post, Chris! Nice job!

    I think Ben Fry’s comment really resonates with my situation – and I imagine it resonates with lots of other evaluators who want to expand their capacity to present smart, concise, and neat interactive data viz but don’t possess all of the requisite knowledge to do so. I know that I can’t develop all of the necessary skills myself – at least not anytime soon. So my questions are these: how do I know what qualities to look for in potential partners for doing this kind of work? Then, how do we efficiently develop enough cross-disciplinary skill to communicate about and produce a high quality and useful product?

    • says

      Thanks Jeff,
      These are big open questions, but it’s my blog so I’ll try to answer :)

      If you do want to do these kinds of things yourself the best way is to make it easy on yourself. If there is an application that makes it easier to create interactive visualizations (there are of course) start there. Don’t jump into programming just because you found a neat visualization code library on the web. Or rather, do, just don’t expect to get to where you want to go quickly and efficiently.

      The qualities you need from your partner really depends on your current capacity and the assistance you need. It’s really easy for people who are not very tech savvy to get into a bad partnership with a programmer who’s just really good at making what they do sound awesome. If the tech speak goes over your head, try to find a third party who speaks their language to get a second opinion.

      Generally though, portfolios and past work are great places to start.

      Over the last 10 years my niche has kind of been in the space between the technical and the project side of things. When problems occur it’s usually when one side or the other starts dictating the course rather than cooperating and trusting the experience of their partner. It’s also helpful to try to at least basically understand the process on the other side. Is something you’re asking for worth the amount of time that it would take to accomplish on the other end? Just because something is do-able and seems like it would be easy to you, does not mean that it is.

  2. Sarah Rand says

    This is a great post! I’m a big fan of data visualization.
    My research group just finished developing an interactive, visual online report on a project we did about computer science education: http://cemse.uchicago.edu/computerscience/OS4CS/. If you are attending AEA, come to our presentation about developing online reports! Saturday October 19 at 9:50am.

    • says

      Thanks Stephanie! I really think you’re going to like my next post (at least it will probably be my next post). And no, I’m not going to give any hints other than it should be fun :)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>