Resources: UPA 2006 Idea Markets
To what degree can facial tracking technology replace human usability specialists?
Activator: Laura Richterich
The Activator's Initial Questions
Facial Tracking Software: Locks-in on 3 muscle-groups of the user's face. Activity in these muscle groups indicates the user is experiencing "struggle" while doing a task. The software does NOT identify facial expressions and translate them into emotions (e.g. smile=happy, etc.). It simply detects the movement of certain facial muscles, and cumulates the amount of activity that indicates negative affect.
Play a Supportive Role to:
...a very limited degree. ...as part of observational data. …attitude measure. …human observations. …to refute the "it was easy!" phenomenon of inaccurate participant self-reporting …only should be used in combination w/ above.
Would need STRONG empirical research to show:
... The technology can be trusted ... The results are reproducible ... The results are accurate as compared with an actual usability specialist
Interim Summary: The technology is especially good as a quantitative measure for developers who aren't as "moved" by human observation (or usability specialists' reports) as are usability specialists.
How would your business partners view this technology?
Its potential for low cost and very quick (albeit very oversimplified) feedback may be overwhelmingly attractive, thus just a little bit dangerous.
Usability specialists would need to only offer it in conjunction with, or as validation for, traditional usability testing. Standalone the technology would not be a good idea, except possibly to identify trouble areas.
The Future is Here:
How would you feel if your laptop's built-in camera assessed you were experiencing difficulty, and contacted the site you were using to push a "social agent" to help you?
Would want to be aware my laptop's camera was watching me first Would want to be aware that this was a possibility Would expect feature to be defaulted "off" and have the ability to turn it "on". It would be freaky.
Interesting Suggestion: Could a study be run simultaneously on hundreds of users on a live site? As they'd use the site, the software could send the "struggle" data (based on facial tracking) back to the study coordinator.