top of page

Learning to Share: The Hacking Open Data #VizForSocialGood Project

I am an only child.

I never had to learn to share when I was younger, and I never had to deal with siblings forcing their opinions and desires on me. I could pick my own activities and indulge my introverted nature to my heart's content. It led to lots of uninterrupted reading, lots of time available to learn how to write code (in BASIC and Pascal; this was the 80s, man), and lots of freedom to create whatever I wanted to create.

While I'd like to think that I've learned how to communicate with real-live people since then, I still have self-selected into career tracks that involve a lot of solo work--writing, editing, coding, designing, and (of course) visualizing data. So what in the world would possess me to sign up for a project that required me to work with another person--and not just another person, but a complete stranger?


I'd missed out on the earlier #VizForSocialGood projects earlier in the year, and I had promised myself months ago that I would make more of an effort to get involved with viz projects that served a greater purpose that simple entertainment. When the Hacking Open Data project came around I decided to go for it, even though the prospect of, somehow, pair-programming a visualization didn't make a whole lot of intuitive sense to me.

I signed up for the Climate track, figuring that everybody in the world actually lives in the world, and thus would have some kind of connection to whatever element of climate data our team decided to work on. I did not, however, pick a specific partner; moreover, I didn't know how long it would take to match me up with anyone, nor did I know when or how I'd be notified about it, nor did I know what the criteria for making matches was. All I knew was that I saw the Tableau Public Twitter account announcing pairings throughout the evening after I signed up, and then on through the next day -- including in the Climate category. Had I done it wrong? Had I messed up the signup process somehow? Was I not good enough to merit a partner? (Or, conversely and with unseemly self-importance, was I so good that they were saving me for just the right person?)

Eventually the anxiety of waiting got too much for me and I passive-aggressively tweeted:

Either my timing was good or my squeaky-wheeling did the trick, because I was soon partnered up with Kim Unger (@WizardOfViz). I didn't know anything about her, and since we weren't ever going to meet in person or be introduced through mutual friends, we were both flying blind in terms of figuring out how we would work together, how we'd divide up work, who had what relative strengths and interests, and so on, and so on.


Since time was of the essence, we both started looking at the data and coming up with ideas about what to do independently (probably before we ever communicated with each other, actually). This was representative of the way we ended up working together, probably because we both felt like it was more important to get the project moving and completed in the time allotted than to stage pitched battles over any one particular thing. One of us would suggest ideas, and the other would pick a favorite out of those ideas; one would find data and sketch out a story, and the other would develop parallel pieces that tell the same story from a different perspective; one would write formulas and calculations and pass them on to the other, who would revise or add on; we would pass the entire workbook back and forth via Dropbox, making sure that we weren't overwriting (or undermining) each other's work, and also making sure that our respective schedules fit in with the time we had to take our turn on building the viz.

One thing that helped the project was that neither of us was particularly precious about the project; while we both cared about the issue we chose (how climate change and the rising sea levels threaten coastal cities and the ecosystem writ large), and we both had a general idea of the story we wanted to tell and the tone to take (I believe the word we used was "morbid"), we both trusted the other to make decisions on any element of the project, at any stage. We were determined that at the end of the process, the final visualization would represent both of us equally and inextricably.


Here is the viz we produced: entitled Underwater, it points out how many people in the United States live within only a few feet of sea level, how soon scientific models predict some coastal cities will be submerged, and how strong the historical evidence is that these models are accurate and indisputable. Our hope was that the raw, historical data, presented cleanly and with clear explanation, would be convincing, and that the interactive elements of the viz would bring home to everyone in the United States just how soon it would affect them personally.


After we published our final viz--relatively early, as it turned out--I messaged Kim:

"I think we did pretty well for not knowing anything about each other's specific skills, interests, working styles, schedules, temperaments, or other qualities that are useful when planning a successful collaboration."

And I still feel that way, but I think she and I got very lucky that our work styles and personalities meshed well enough that we were able to accomplish this kind of remote collaboration successfully. Vizzing as a duo is very difficult; we solved the problem by taking turns and passing the workbook back and forth, while also working in secondary, scratchpad-type workbooks when it wasn't "our turn" so that we could work out some complicated calculations or design challenges. Communication was done entirely through Twitter; maybe we could have done better by trying to Skype or use Google Hangouts to get a better sense of one another's personalities, but it didn't seem to matter. Since we didn't know who was "better" than the other at any particular element of data visualization, we were always willing to defer to the other if an issue arose (although this was rare, if ever). We only found out very late in our project that there was a Slack channel set up for participants; I never joined and I don't think Kim did either.


Lessons learned for next time: I would rather have a partner I knew or who I knew more about, so that it would be easier to effectively divide tasks among each teammate's preferences and strengths. Kim and I never had that conversation, maybe because we didn't think we had time to negotiate before getting underway. I know some other teams divided work into "ok, you find the data and do the ETL, and I'll do the visualization," which isn't how I'd want to work if my partner and I were matched up out of the dataviz community. If we were going to work that way, I would have selected/preferred a different partner, maybe someone with limited viz skills but killer Python and Google-Fu.

I also heard cases of partners struggling with being on different schedules or in different time zones; or with one partner being less committed to the project than the other (endemic to many group projects, I think). Again, I was lucky that Kim and I were equally committed to doing this, and probably both felt a responsibility to one another, as well as to the overall Hacking Open Data project, to invest a significant amount of energy and thought into our viz, so that we'd end up with something to be proud of and that would be useful in spreading the message of how severely climate change will affect us in our lifetimes.

I would participate again, in just about any #VizForSocialGood project, because I feel that it is important to use whatever skills we all possess to contribute positively to the world. Doing a group visualization would make me a little more concerned; not because my experience was bad, since it wasn't; rather, because it adds so many challenging elements that others found daunting and hard (or impossible) to overcome, I'd be worried about not being so fortunate next time around.


Recent Posts

See All
bottom of page