Table 4.

A Description of the Spatiotemporal Properties Along With Their Corresponding Ranges

Category	Property Name	Description	Range
Spatial	1. PED: Amplitude	Describes the most frequent positional error observed. Higher values of amplitude for a mean of zero indicates better performance.	[0 1]
	2. PED: Mean	Describes the spatial offset. Values (in visual degrees) closer to zero indicate better performance.	[0 ∞]
	3. PED: Standard Deviation	Describes the spatial uncertainty: the spread of the positional deviations. Lower values indicate better performance.	[0 ∞]
	4. PED: adjusted R2	Describes how close the positional error distribution resembles a Gaussian distribution. Values closer to 1 indicate better performance.	≤1
Temporal	5. Average Velocity CCG: Amplitude	Shows the maximum correlation between the stimuli and gaze velocities. Higher values indicate better performance.	[−1 1]
	6. Average Velocity CCG: Mean	Describes the temporal lag between stimuli and gaze velocities (in ms). Lower values indicate better performance.	[0 ∞]
	7. Average Velocity CCG: Standard Deviation	Describes the temporal uncertainty: the time window (in ms) in which the observer is uncertain in their ability to track the stimulus. Lower values indicate better performance.	[0 ∞]
	8. Average Velocity CCG: Adjusted R-squared	Describes how close the temporal tracking performance resembles a Gaussian distribution. Values closer to 1 indicate better performance.	≤1
Integrated	9. Observation noise variance	Describes the noise internal to the observer Sensory noise estimated by measuring the variance of the observational noise using a flipped Kalman filter. Lower values indicate better STP.	[0 ∞]
	10. Similarity	Cosine similarity between gaze and stimulus vectors of positions. Higher values indicate better STP.	[0 1]

CCG, cross-correlogram; PED, positional error distribution.