User Centered Design — Don’t Ignore or Believe, but be the Midwife or the one who does some Artificial Insemination

«We’re the other kind,» said Granny. «We’re the kind that gives people what they know they really need, not what we think they ought to want.»
«Witches Abroad» by Terry Pratchett

I’m still advising students when it comes to product development and evaluation (computer science & media, or more generally human-computer interaction). I love the task, but one thing that still strikes me is how difficult it is for some students to actually develop ideas. I mean, the first step in the human centered design process is an analysis of the use context. The contact with the actual users should give them some ideas what they can do to develop an application that is efficient, effective, easy to learn, and provides a good experience.

However, there are at least two problems that frequently occur: Either ignoring the target group or uncritically believing what they say.

Ignoring the target group: The students don’t ask and listen (or observe) — they just follow their own first ideas without challenging them. And sure, this is easy and you can even do some pseudo-surveys and interviews later to confirm the first ideas (confirmation bias is a thing, just go looking for examples and you see it’s true ;-)). But if they don’t listen, they fall prey to the avoidable tragedy of engineering — the breaking of a beautiful artifact on a neglected reality. I try to deal with this problem by having the students sketch out of their ideas at the very beginning of the project (by hand so the investment is still low!). They have these ideas anyway and now they are at least out in the open. I make a copy of those sketches and tell them: If by the end of the project you don’t have something different, you didn’t do your job properly. (And yeah, there can be cases when the first sketches are a “perfect fit”, but I think these are extremely rare — and almost never occur if they are not also members of the (in this case very homogeneous) target group.

Uncritically believing what they say: Sometimes students listen too uncritically, esp. when the target group is not savvy when it comes to the development of media and computing technology. The target group rarely has the background, interest, and did the necessary (prior) effort to develop what they actually need from the product. Latest example was the assertion that the target group does not need statistical information in the App, given they are looking for relationships between specific behaviors and consequences of these behaviors — and not simply collecting some observations. … Yeah, well, one major use of statistics is the discovery of relationships, esp. when human have difficulties finding them! And as someone who really loves data — it’s a geek thing — this assertion actually hurts. Instead of uncritically accepting these assertions, some further analyses and questions would have been helpful. For example, using sample data to show what statistics are capable of — or providing some visualizations of possible relationships. Not with an “I know what you want, I’ll show you how you’re wrong” attitude, but rather with a “Let’s just make sure this isn’t something that could make your life a bit easier.” The result might still be the same, but with less uncertainty.

I think it’s very difficult to strike the right balance between ignoring the target group by just “knowing” what they (should) want and uncritically believing what they say. I don’t know what’s worse. In the first case, you get an app that shatters on implementation, in the later you get a mediocre solution that is far below what is possible. I guess that later has at least the advantage that it works — somewhat.

For addressing the first problem, I like the metaphor of the midwife. Similar to a midwife, you can help, but you can only help. However, with the second problem of not being critical enough, I think sometimes it’s necessary to do some artificial insemination. To plant some ideas and see whether they grow. And not to stretch the metaphor too far, but, yeah, the material should still come from the target group.

Anyway, just some thoughts … still thinking about the topic.

(And dang, I need to work on my metaphors …)

1 Comment

  1. I had a long talk with Henrik about this early prototype thing last week (and like two months ago (?)) and our positions were quite similar to yours, although our intentions were quite different at first. We thought that if you have something in your mind on how your “product” will look like at the end, just put it out into the world and test it with your pontential user group(s) and gain information about your prototype from the early beginnings (in terms of teaching the user centered design process, hopefully this prototype has many flaws or at least the confirmation bias could be seen and discussed).
    I would say it does not matter if you have the picture just in your mind or on paper/screen/whatever but the difference is, you can observe how people can interact with it if you show it to them which is a huge plus.

    Besides, I am just reading “Just Enough Research” by Erica Hall (which is a fantastic book, especially if you are not into research). And I think HCI and media informatics students could learn a lot from it.

Comments are closed.