Transitions in construction

Transitions in construction

Whatever you thought, think again.

Pull on this.

Step-changePosted by Dan Engström Thu, January 05, 2012 01:24:21
Let’s assume you’re done with a research project or a development of some sort. You now have a data set for your colleagues to understand. It might be the interaction between actions in a product development process, the number of sick days for employees in different business units of your company or the connections between your key performance metrics. It might even be the connections between the characters in Les Miserables, which is what they did in the graph below.
It doesn’t look like much, I know. And just showing a static version of it doesn’t do it justice. It’s interactive, you see. You get to manipulate it. It’s so cool and opens so many opportunities. Since my blog medium is very basic, I’d like to you to do something for me. I want you to leave my blog for a minute or two and play around with it (link here). Let the cursor point to a spot of colour and see which character it is. Click on a spot of colour with only a few strings attached and pull. Then click on one with more strings attached and pull on that. Go do that and then read the rest of this post. Knock yourself out. I’ll wait for you here.
OK. Have you done it? No cheating. It’ll only take you a minute. If you work where pure evil decides what you can and cannot view on the web, play with it at home. Seriously. This is amazing stuff. Go do it. In the meantime, I’d be happy to play you some elevator music.
Now have you done it? Good. Cool, wasn’t it?

It’s a force-directed graph. It shows you the relationship between a set of points and how manipulating (yes, pulling on) a certain one will affect the rest. Pulling on one with many connections will affect the whole data set. Pulling on one with only one connection will do very little. The width of the possible uses for this I can only begin to imagine.

This force-directed graph template is just one of many data driven documents available for download at D3.js, a small, free JavaScript library for this sort of graphs. D3.js is meant for people like you and I to find the perfect templates to use as pieces in our own jigsaw puzzle. You pick a suitable template for a data-driven graph, let it read your data from a source file (which can interact on command with your original research results data if you like) and then you get an image that was meant to be manhandled. It was designed to be used by an active group, not looked at by the lone ranger.

At the page for the Les Mierable graph, you’ll find the source file to be a certain miserables.json file. As of yet, I have no idea what format that is in or how to fill it with my own information, but I am sure going to find out.

So imagine a workshop where you use this to show the results of your project. Will your audience be interested? Will you make them curious? Will you make them write down the graph URL and go back to their office in order to be the cool guy that shows it off? Hell yes.

These graphs give us the tools to take responsibility for the implementation of our project results. They let us quit being consultant researchers that take the jobs thrown at us, that send reports and move on. They allow researchers to take active part in business. That’s not a bad thing.

And here’s the key. When we have collected some form of data, which is a key ingredient in R&D, we look at it from different angles, write the text and then develop suitable illustrations to suit the text and data. Very respectable, but it doesn’t make anything happen in the minds of business people.
I suggest industrial researchers turn this established procedure upside down and do the exact opposite – develop the project from the graph. Work on what can be illustrated in a meeting room, instead of trying to make everything you’ve worked on come across. By all means write a report too. Just remember that if we have to present a report for the people in the room to be able to decide whether to take any sort of action, we’ve likely failed. Our R&D project is dead in the water and a good deal of the investment wasted.

But then again, maybe you don’t write shelf warmers. Your reports are so frequently used by others that you want to stick to sending people a 37-page pdf document in the mail beforehand.

No. I thought as much.

Well, if you don’t yet believe in the power of the image as primary medium for collecting and examining data, go to this post at the blog and have a look around. Check out, say, how the world is feeling right now through and you’ll be ready to write a blog post just like this one.

Credits: Force-directed graph from d3.js. Images from Flickr Creative Commons: bored by roy costelly, Muzak dial by Ryan Harvey and Turn The Whole Thing Upside Down by Melissa Gray.