eyetracking-methodology.pdf

(16187 KB) Pobierz
Eyetracking Methodology
How to Conduct and Evaluate Usability Studies Using Eyetrackin g
Kara Pernice and Jakob Nielsen
August 2009
WWW.NNGROUP.COM
48105 WARM SPRINGS BLVD., FREMONT CA 94539–7498 USA
COPYRIGHT © 2009 NIELSEN NORMAN GROUP, ALL RIGHTS RESERVED.
To get your own copy, download from: http://www.useit.com/eyetracking/methodology
829489208.033.png 829489208.034.png 829489208.035.png 829489208.036.png 829489208.001.png 829489208.002.png 829489208.003.png 829489208.004.png
This page is left blank for double-sided printing.
2
INFO@NNGROUP.COM
Executive Summary
829489208.005.png 829489208.006.png 829489208.007.png 829489208.008.png 829489208.009.png 829489208.010.png
Contents
Executive Summary .......................................................................5
Should You Use Eyetracking in Your Usability Studies? .........................6
Eyetracking Method Tips ...............................................................7
Recruiting Users for Eyetracking Studies..................................... 13
Communicate to the Participant During the Screening Interview........13
How to Say It ......................................................................................14
What to Ask the User ..........................................................................15
Telephone is Easier than E-mail When Recruiting for Eyetracking
Studies ................................................................................................18
How Many Users to Include in a Study ........................................ 19
The Number of Test Participants Needed is Dictated by Desired
Output.................................................................................................19
Variablity in Heatmaps Depending on the Number of Users ................20
The Slowly-Diminishing Returns Curve ...............................................48
R 2 Explained ........................................................................................49
A Better Deliverable Than Heatmaps...................................................51
Eyetracking Studies in a Lab........................................................ 53
Eyetracking Lab Set-up Notes .............................................................53
Think Aloud Method..................................................................... 57
Benefits to Think Aloud in a Non-Eyetracking Study............................57
Drawbacks to using The think aloud Method in a Non-Eyetracking
Study...................................................................................................60
Drawbacks to Using The think aloud Method in an Eyetracking Study.60
Surveying is Common During Think Aloud...........................................67
Beginning the Session and Calibrating the User’s Eyes................ 72
What to Say to the User When Beginning the Sessions .......................72
© NIELSEN NORMAN GROUP
WWW.NNGROUP.COM
3
829489208.011.png 829489208.012.png 829489208.013.png 829489208.014.png 829489208.015.png 829489208.016.png 829489208.017.png 829489208.018.png
Impossible Eyetracker Calibration............................................... 87
Cannot be amended: These culprits really cannot be changed to
make calibration possible....................................................................87
Can be amended: These culprits usually can be changed to make
calibration possible .............................................................................88
Facilitating Test Sessions .......................................................... 104
Steps for making eyetracking test facilitation run smoothly .............104
Example of Facilitator’s Protocol Sheet .............................................110
Saving Files .......................................................................................111
Retrospective ....................................................................................112
Using the Right Eyetracking Analysis Tools ............................... 116
Gaze Replays.....................................................................................116
Gazeplots ..........................................................................................117
Heatmaps ..........................................................................................118
The Areas of Interest or LookZone Feature .......................................138
Analyzing Video and Animation .........................................................139
Using the Eyetracking Analysis Tools in the Best Way.......................143
Tasks Discussion ....................................................................... 145
The Scenic View: Show One Page and Interview Method .................145
Task-Based Testing ...........................................................................148
Variety in Tasks.................................................................................149
Should You Run Your Own Eyetracking Research Study?........... 152
Why Nielsen Norman Group Invests In Eyetracking..........................153
Technology-Related Notes......................................................... 155
About the Authors ..................................................................... 157
Acknowledgements ................................................................... 159
4
INFO@NNGROUP.COM
Executive Summary
829489208.019.png 829489208.020.png 829489208.021.png 829489208.022.png 829489208.023.png 829489208.024.png
Executive Summary
In traditional usability testing without eyetracking there are many cues our test
users give us that help us evaluate designs, such as the time a person spends on a
page, the action he takes there, what he reads aloud, what he hovers the mouse
over, whether he smiles or grimaces, and what he comments on. These user
behaviors comprise, in our opinion, the most important aspects of learning about
usability. But eyetracking (ET) adds another dimension. Watching what people are
looking at adds a level of interest to usability studies. Following the eye at work
makes you feel as though you are in the user’s head, thinking with him. If nothing
else, the ET studies are more interesting and hold observers’ attention more easily.
Eyetracking technology also helps test facilitators avoid a major blunder: interrupting
the user when he is quiet for a long time. This is a huge benefit since interrupting
the participant too much or inappropriately is one of the greatest challenges for even
seasoned usability professionals.
In addition to helping to avoid a major facilitation gaffe, eyetracking technology
enables us to understand more than clicks and pauses. We learn finer points about
what draws attention and why. And by studying what people look at and do not look
at, we are able to glean more detailed insights about how the Web user works his or
her way through usable and unusable designs. Behaviors such as exhaustive review
(where people look repeatedly at areas that seem as though they’d be helpful but
are not) and selective disregard (where people purposefully tune out areas of the
website at given times) are very apparent when watching the user’s eye at work. We
see which interface items are miscues (erroneously calling attention) and which
areas trick users and are thus dropped like a hot potato. (All of these and more
behaviors are described in our book titled Eyetracking Web Usability , Nielsen and
Pernice, New Riders Publishing 2009.)
Conducting valid usability research calls for good planning and experience. But
conducting valid usability eyetracking research is on another plane, adding to the
preparation, facilitation, and analysis. Eyetracking technology takes a regular
usability test and makes it more difficult, more time-consuming, and more
expensive. Conducting or analyzing results of a research study erroneously is worse
than not doing any research at all. Unfortunately, ET technology makes it very easy
to get misleading results where you believe something that simply isn’t true about
your design.
In this report we discuss what we have learned about how to plan, conduct, and
analyze an eyetracking study that will produce sound results that you can trust and
use to improve your website. While we see many benefits to eyetracking research,
we also know that good usability studies do not call for high-tech solutions. Some of
the best, iterative tests are done with paper and pencil only, because the faster you
can run a study the sooner you can adjust the direction of your design project.
METHODOL OGY, NOT FINDINGS
This report does not present any of the findings from our ET projects—it’s purely
about methodology : giving advice for how you should run your own ET studies,
should you decide to do so.
© NIELSEN NORMAN GROUP
WWW.NNGROUP.COM
5
829489208.025.png 829489208.026.png 829489208.027.png 829489208.028.png 829489208.029.png 829489208.030.png 829489208.031.png 829489208.032.png
Zgłoś jeśli naruszono regulamin