try it now!
try it now!
You can do this.

1. Get the reading environment set.

If your work comes on CD-ROM, have the disc out (might as well see whether the reader can figure out how to load it). If your work appears in a browser, don't forget to clear its history so the visited links won't give any cues.

If you can, approximate the environment in which the reader usually reads. If not, at least get the reader comfortable. Refreshments help.

2. Put aside your authorial creative role, and move into an enfolding accepting open deliriously-grateful-for-anything-the-reader-says data gathering role.

If you find you just can't do this, but are still interested in feedback, you can have a friend adminster the review for you.

Note that the administrator role is friendly but opaque, giving almost no information, just open-ended reassurance. The best usability administrator I know, Alison Fields, answers questions like a particularly frustrating therapist:

"What does that do?"

"What do you want it to do?"

"Where's the <feature>?"

"Where do you want it to be?"

Or Alison says,

"Rather than me telling you, we're much more interested in your thoughts and opinions."

3. Get something to write on, or arrange for the session to be recorded.

Note that you must ask the reader's permission for any recording. When getting permission, explain the purpose of the review and how the results will be used.

A great method is to have someone else lead the session, while the author sits there, taking notes but not talking. At the end, the reader and the author can talk freely. (At this point most authors will have a lot to say.)

4. Greet the reader, who has just arrived. Use the magic incantations to lull the reader into the right mood, avoiding words of judgment or even discussion.

Humans can really only do one thing at a time. After only about five minutes of awkwardness, if you've set the mood correctly, most readers will relax into the process and be almost completely natural. Remembering to be tense, when they're engrossed in your work, usually proves to be too much mental overhead.

5. Shut up while the reader reads, unless you have to (1) urge the reader to think out loud or (2) ask for more data.

To ask for data about links, you might use UIE's link focus questions (included in the incantations below). They sound cumbersome but after only a few links you and the reader will get into a rhythm and together discover some astounding stuff.

6. Don't twitch or gasp no matter what happens. Also try to keep the author from twitching or interrupting. (Especially if that's you.)

>>> Avoid any words that might destroy the magic.

7. Let the reader stop when s/he gets tired or feels the reading is done.

8. Now you can debrief and discuss, all together, without worrying about muddying the reader's all-important "fresh" responses.

You did it!



My Passionately Held Opinions About Using "Findings"

Selection of good readers (="readers who are a lot like, if not identical to, the people you intended the work for") is the #1 factor in a useful review.

No matter how a text (or system) works, readers will make their own utterly amazing mental models of how it works, and act according to them. We humans are obsessed with discerning meaning in chaos, online and elsewhere, and tend to discern it whether it's there or not.

Reviewers are often too polite. Don't listen to what they say: watch what they do. Performance-based tests, that describe mostly what the reader did, are the kind I prefer.

Reviewers are not writer/designers. A usability review is excellent for finding where problems lie, but then the problems have to be turned back to the writer for solution (or at least another crack at it). A common occurrence: four readers come to grief at the same point. Each one has a different explanation of why, and a different recommendation for how to remedy the problem. Those can be reported to the writer if there is interest, but the most important thing to report is "four readers all crashed at node X," and it's up to the writer to take it from there.

The Magic Incantations   

[You say these.]

We are [or: the author is] incredibly grateful for your time. You will be providing a valuable kind of feedback that will help make the work a more [intense, wonderful, your adjective here...] experience for future readers.

In everything you are about to experience, there is no right answer. Whatever you think, is right. [People who have been over-conditioned by some types of schooling will not believe this and will try to "trick" the right answer out of you anyway.]

I'm going to sit with you during the review, and I might suggest areas to explore. However, I'm not usually able to answer questions about the text. This is not to be cold-hearted, but because because what's important is what you think, where you are looking for the next part of the [story or other noun here].

Don't worry if the web site [or program] crashes. We'll just start over. [Or: we have plenty of electrons. Or a joke of your choice]

As you read, please "think out loud" - tell us what's going through your mind. At some points we may ask you to "surf" a bit more slowly than you would normally, so we can talk about your reading choices. For example, when you're about to click on a link, just hover over it with the mouse and we'll talk briefly before you click and also briefly afterwards. The questions will always be the same:

(before)

   Why are you clicking?
   What do you think this link will give you?

(after)
   Is this what you expected? Why or why not?
   (optional for Deena - grin) After clicking this link, do you feel closer, or farther, from "closure" within this piece?

When you feel you have read enough, say, "I'm done."

[No matter what happens, repeat at intervals,]

"This is great feedback!"

"This is just what we were curious to know."

"This is great / just wonderful / very helpful." (etc.)

Words of Magic-Destroying Power    

[You don't say these.]

I believe (at least based on my experience in the U.S.) that many of us are trained by school to look always for the "right answer." A nervous mental rigidity kicks in whenever we encounter a situation that's at all like school.

I would recommend that people sitting with a reader zealously eliminate all of the following words from their vocabularies:

test - no, it's a "review" or a "usability session"

"It's supposed to..."

"You should have..."

"It's better when you..."

"That's wrong, because..."

At this stage, don't disagree or tell them what you intended. Whatever that might have been, they're reading now. (People with post-whatever theory expertise might be less likely to tell readers they have no right to their opinions, than my software colleagues typically are, so perhaps I'm worrying about nothing.)

Old Links

I've written about usability before, for HTWW events; each time I try to do a better job! But here are the old versions, for reference:

1998 version
1998 afterpaper - some usability resources
1999 version - off topic, but footnotes have a reader expectation rant

what I've learned so far   ·   try it now!   ·   list o' resources   ·   excerpts from the March 2000 chat   ·   invitation to chat   ·   a blank page




Copyright 2000 by Julianne Chatelain - last updated 29 March 2000
See also professional - personal - home