Calibration for Eye Tracking
To collect accurate eye tracking data, a short calibration must be done with the participant. This calibration usually takes less than 1 minute. Participants are asked to look at some dots and click them. Point and click is by far the most accurate method of calibration as other methods that don’t rely on interaction are unable to achieve the same level of accuracy. In some special circumstances we may be able to calibrate without interaction, if the content meets certain criteria. Contact us for more information on this.
Recommended Calibration Points
Stim Admin allows calibration points to be shared between Stims reducing the overall effort required.
The number of calibration points should be determined by the length of the viewing session and/or number of stimuli. We recommend 30 calibration points (15 before and 15 after), plus an additional 5 between stimuli.
Fewer calibration points can be used at the user’s discretion however lower numbers will reduce overall accuracy and how long the calibration stays accurate. Projects have used calibration points as low as 15 (5 before 10 after) and achieved good results.
Participant motivation during Stim viewing also affects the amount of calibration needed. If participants are not interested in the content, their attention will wander and the calibration will become invalid more quickly.
NOTE: In StimEngine calibration points are calculated on a factor of 5, meaning 1 is equal to 5 calibration points.
Sequence and time of stims = Recommended calibration [number of points]
(30s) + (30s) =  (30s) +  (30s) 
(30s) + (1m) =  (30s) +  (1m) 
(30s) + (2m) =  (30s) +  (2m) 
(2m) + (30s) =  (2m) +  (30s) 
It is possible to share calibration between Stims in separate Stim Admin viewing sessions, e.g. with survey questions in between Stims. The benefit of sharing the calibration is the overall reduction in participant effort. If this is something you are looking at implementing please contact us and we can walk you through the technical requirements in more detail.