Eye Tracking—Stuff that makes you crazy… >8-/

When I read in the raw data, I check each point to check if it is valid and flag it if there is a problem.  Sometimes the tracker will lose the eyes, so during any period of measurement there is a probability that some of the points in the record will be invalid.   What I failed to do was remember that I had done that and take the validity of the point into account in generating graphs and doing analyses.  

Luckily, that generally means that the graphs and statistics will improve since it appears that the tracker reports a gaze point of (0.00, 0.00) when it loses the eyes.  While it is possible for a participant to look exactly at 0.00,0.00 it is virtually impossible for the tracker to report a true gaze point of 0.00.  The position of the pupil is determined, then mapped to the screen pixels via what is likely to be some complicated trig, and the mapping is never going to be perfect. There will always be a bit of error.  Couple that with the simple precision errors that can accumulate in “double” datatypes it is almost guaranteed that a true report of exactly 0.00, 0.00 will never happen.  So I interpret such reports as invalid data. 

Now to re-export the data and analyze again…

This entry was posted in Eye Tracking. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s