[Sigia-l] eye tracking?

Christopher Fahey chris.fahey at behaviordesign.com
Mon Jan 9 21:07:40 EST 2006


Stewart Dean wrote:
> I personaly don't feel eye tracking is of any real value. It
> shows what users are looking at but not really what they
> are thinking or what they are seeing when they look at
> something.

I was thinking the same thing. Eyetracking seems like a tool that would be
most useful for a group of people who either have no graphic design sense at
all, or who at some fundamental level don't trust graphic designers (or
traditional usability testing). As Stewart Dean excellently points out, just
because a user notices a page element five seconds sooner or for 25% longer
than another element else doesn't reveal much about that user's
*understanding* of the elements on the screen -- in fact, such data may lead
to radical and arbitrary misinterpretations of the interface's performance. 

And all that stuff about chinrests, calibration, dim lights... To me that
sounds like a surefire way to make a test subject incredibly nervous and
self-conscious! 

A simple, traditional usability test using plain old-fashioned talking will
more successfully uncover the real *meaning* of the user's experience,
despite wherever their eyeballs are pointing. If you need to learn whether
or not users are noticing critical page element X, simply ask a bunch of
them (a) What did you notice first? and (b) Did you notice X? You'll get
amazing insights.

Still... If I were Google, and I were redesigning the Google home page to
include a banner ad, and I needed to prove to potential advertisers that the
banner ad would be seen instantly, I'd probably want to spend a few bucks
eyetracking a couple of designs to pick the one that would empirically hit
more optic nerves than the rest.

Or, if I were designing an air traffic control system and I needed to place
a critical alert pop-up box in just the right place on the screen so that
controllers would see it before a plane crashes, eyetracking testing might
help pick that final alertbox position.

But for projects with less than a few billion bucks on the line, with no
human lives or split seconds at stake, and with competant graphic designers
on the team, I am pretty skeptical about the usefulness of eyetracking.

I'm perfectly happy and eager to hear rebuttals to this skepticism, as I
have zero actual experience with the technology and techniques.

-Cf

Christopher Fahey
____________________________
Behavior
http://www.behaviordesign.com








More information about the Sigia-l mailing list