Beki Grinter

HCI and Computational Thinking

In computer science, empirical, HCI, research on February 24, 2011 at 10:42 am

Mark and I both received a pointer to the iConference 2011 website, one that would trigger Mark to write a post about the differences between HCI and Computational Thinking.

If I grasp his argument correctly, it is the surfacing of a tension.

Some visions of HCI-oriented design have explicitly argued for making the machine disappear. By contrast, Computational Thinking requires that people understand the machine. It’s presence is not diminished, it does not disappear, but rather it becomes more clearly surfaced, so that its capabilities and its limitations are understood.

I have some thoughts…

First, the HCI community has discussed the difficulties of having a vision of disappearing as the design goal. One really nice contrast was “Unremarkable Computing” which suggested that the artifact was not designed to disappear perse, but that the appropriation processes end-users go through work the machine into practice. Appropriation is about exploring, understanding, accepting, and then finding a means by which the machine comes to be useful despite its limits. Matthew Chalmers work on seamful design asks can you exploit the seams, bring them into the design itself, and use that to shape engagement. Susan Wyche’s work on Extraordinary Computing suggests that there are times that the design might want to emphasize the machine or application, to make it a part of the experience very explicitly, rather than hidden away in an experience where the machine is forgotten.

Second, I think some of this tension turns on the task at hand. There are times when it makes sense to make the limits explicit and times when those limits need reduction. It depends what is going on. Mark’s picked some great examples of occasions when understanding the limits is really crucial. For people tasked to learn and educate (formally, informally and so forth). Another area I think that this applies to is in IT policy for corporations. I read in a Harvard Business Review that over 50% of all CRM systems fail in deployment, largely because it is assumed that it is a system that is being adopted, not a system that implements a management philosophy that is being adopted. Applications that claim to help an organization manage its processes bring a set of assumptions about what’s appropriate implemented in software and its the failure to learn what those are and whether they interact with the current processes implemented that is the root of many problems. Perhaps that’s less about the machine itself, but it’s a type of learning that only arises as a result of the limitations of the machine.

But, there are times when we want to get something done and the machine and quite a bit more could usefully disappear. Like, for example, in my home network. Mostly I want to use the network, whether it’s streaming stuff or just checking my email. But when it breaks down, or when I attempt to add something into the mix, then I have to budget time for a painful learning curve.

For example, when we decided to switch to U-Verse we wanted to buy a switch. We did not want a managed one, because we didn’t want the configuration burden. So, we needed an unmanaged switch, but we did need one that did IGMP v3 snooping. Why? Well as I learnt we were going to have our television and data go through the same switch. TV over IP 🙂 takes up a lot of bandwidth, and IGMP v3 snooping prevents the switch from sending that (multicast) traffic to any port that is not listening to it. If you don’t have this, what can happen is that if you have, say, a wireless network connected into a port it is now receiving all your television and that really slows down its performance. At this point you are wondering what’s so bad about a managed switch, well that’s a whole other post. Now the person in me who received some Computer Science degrees feels rather good that I’ve learnt this (although I would have elected to learn it in class rather than as part of a weekend spent trying to get to the bottom of this), but this I think might be an example of an occasion where learning has gone too far. This sort of thing, if it could, could usefully disappear in my mind. Leigh Star’s work on infrastructure, surfacing it for inspection, but highlighting its invisibility in use, inspires.

So, what I am saying is that I see a relationship between the undertaking and the degree of Computational Thinking required.

Third, at least in my own work, I’ve always seen HCI as mediating between the human experience and computational solutions. Applications are one form of computational solution. But there are a world of others. I spent a lot of time applying HCI to software development processes. I got there because sometimes it’s easier to change people than change technology. That kind of design starts in a different place, and has no ambition to make a technology disappear through the delivery of an application. I think there are other opportunities being taken by other researchers that also get at a version of HCI that should reduce the importance of the application and instead emphasize the importance of the mediation.

  1. […] students at a slightly higher level of abstraction. As Beki Grinter points out in her interesting response to Mark’s post, sometimes we really do want to hide some details from users. Then the question becomes, what […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: