Jonathan's Corner
(Search & Sitemap)
> Orthodox Books and More >
Technology >
Usability for Hackers
Previous
1
2
3
4
5
Next
Printer-Friendly Version
If you're reading this text, there's a good chance that you are already halfway to being an anthropologist. Note: for the purposes of this chapter, 'anthropology' is used to refer to cultural anthropology. Other anthropological disciplines exist, but it is cultural anthropology and its techniques which are most directly relevant here.
How could an author know that you are probably at least half an anthropologist? Let's turn the question around, and suppose you are a Python hacker. Why are you reading this article? Visual Basic .NET has enormous marketing muscle behind it, possibly eclipsing the marketing budgets for all open source technologies put together. Guido van Rossum holds a dim view of marketing, as does much of the Python community. Monster.com lists three thousand Visual Basic positions, almost five thousand .NET positions, but only one thousand Python positions. Why are you reading a "usability for hackers" article when you could be reading a title like Completely Master Visual Basic in Thirty Seconds or Less?
You are probably a hacker. It does not matter if you were mortified when you found out the preferred JavaScript technique to create an object with fields that aren't globally accessible variables, or if you wince when you hear of a devious way to get technology to do things that shouldn't be possible, or if you have no desire to be considered a 133t hax0r. You're probably a hacker. The classic "How to Become a Hacker" for the most part outlines things that have a very obvious relationship to being a hacker: attitudes towards technical problem solving, or learning an open source Unix, learning to program and contribute to the web, and so on and so forth. Towards the end there is a particularly interesting section because on the surface it looks completely beside the point. The section is titled "Points for Style," and mentions learning to write well, reading in science fiction, training in martial arts, meditation, music (preferably obscure), and wordplay. Other things could be added: avoiding mainstream TV or having arcane hobbies and interests, for instance, so that in a social context hackers may ask each other questions about obscure hobbies as a rough social equivalent to, "What's your favorite TV show?"
Not that any of these is necessary to be a hacker, but together these common trends point to a personality profile that can learn the anthropological style of observation relevant to usability work much more easily than the general public, or even Joe Professional Programmer who regards learning new technologies as a necessary evil rather than a joy, works in Visual Basic .NET after being swayed by advertising, goes home and watches TV after work, has probably never heard of ThinkGeek, and would probably rather do gift shopping at Walmart even if he does know of ThinkGeek.
All of this is to say that the culture surrounding you is not like water to a fish. It is a basic fact of life that you don't automatically share the perspective of others. Cross-cultural experience or ethnic minority status may accentuate this, but this is true even if you're not (regarded as) a minority. And this kind of experience provides a very good foundation for anthropological ways of understanding exactly how you are not a user and users don't think like you.
One basic challenge for organizing a site's information architecture is the taxonomy, or way of breaking things down. If one is asked what an example of a good taxonomy, one example of a taxonomy par excellence is the biological taxonomy that organizes all the way from kingdoms down to species or subspecies and varieties. And indeed that is one kind of taxonomy, but it is not the only possibility. If one is asked to break down a list of a fork, spoon, plate, bowl, soup, and macaroni and cheese, one obvious way is to put the fork and spoon together as cutlery, the plate and bowl together as dishware, and the soup and macaroni and cheese together as food. But this is not the only basic way, and it can make sense to put the fork, plate, and macaroni and cheese together as representing one complete option, and the spoon, bowl, and soup together as representing another basic option. Stores and websites that have adopted the latter approach, such as a gardening store or website that organizes its products according to the type of garden a customer is trying to make and what the customer is trying to do, see a significant increase in sales. Even biology could use other complementary technologies: a taxonomy that classified organisms according to both ecosystems and their roles within their ecosystems and ecological subsystems could say something very valuable that the eighteenth century classification wouldn't.
In terms of websites, an information architecture that corresponds to the organization's org chart is never a helpful choice. Even when we are talking about an intranet intended only for organizational insiders, one section or subsite for each department is not the right choice: one better option would be to support workflow and design around the tasks that employees will be doing with the intranet.
What is the best information architecture? That's not a question to answer by looking something in a book or even thinking it out; it is something that we should work out based on what we observe doing research, even if we also read and need to do a bit of thinking. And this is the best practice across the board for usability.
One valuable exercise to help guide information architecture design is called card sorting. In this exercise, we get a stack of index cards, perhaps 3x5", and write the individual names of different pieces of functionality the website should offer, trying to name things neutrally so that the names do not have common terms suggesting how certain parts belong together. Then we shuffle and lay out the cards, and individually ask subjects (people who will participate in an experiment and who are not insiders, whether employees of your organization for an external website, or information technology professionals) to organize them so that cards that belong together are put in the same stack.
Then we note which cards have been placed together, thank the subject, and move on to the next person.
On looking through the notes, we may see a few things. First, not all people think the same. We will likely see some breakdowns that are very similar, but there will likely be two or more breakdowns as fundamentally divergent as our breakdowns of the fork, spoon, plate, bowl, soup, and macaroni and cheese. Second, there will probably be a breakdown that simply catches us off guard. And this is good; it means the exercise is working.
After doing this, we can go about looking for a preferably standard information architecture that will gracefully serve the major ways we observed of breaking things down.
With an eye to how to best approach observation, we would like to take a moment to talk about Coca-Cola's blunder with "New Coke" and explain why focus groups, bringing in a group of people and asking them what they want, are deprecated as a recipe to make products that look good on paper but don't wear well in normal use. For those of you who don't remember the uproar some years back, the Coca-Cola company announced that it was switching to a new and improved formula, and there was massive public outlash from people who wanted the old Coke back. (Now the company sells both the old formula as Coke Classic and the new formula as Coke II, and Coke Classic is vastly more popular.)
Why would the Coca-Cola company announce it was terminating its cash cow? The answer is that it did naïve marketing research, ran taste tests, and asked members of the public which they would choose: the formula today sold as Coke Classic, or the formula today sold as Coke II. The rather clear answer from the taste tests was that people said they would rather have the new formula, and it was a clear enough answer that it looked like a sensible course of action to simply drop the second-best formula. It wasn't until everybody could see that the Coca-Cola company had given itself a PR black eye that the company woke up to a baseline observation in anthropology: the horse's mouth is a vastly overrated source of information. Most anthropological observation, including the kinds relevant to usability, are about paying close attention to what people do, and not be too distracted by their good faith efforts to explain things that are very hard to get right.
Jonathan's Corner
(Search & Sitemap)
> Orthodox Books and More >
Technology >
Usability for Hackers
Previous
1
2
3
4
5
Next
Printer-Friendly Version