FAQ: What determines whether a validation is "GOOD", "FAIR" or "POOR"?

[Sam]

The system uses the following criteria, measured in degrees of visual angle (°), to rate the quality of a calibration.

• GOOD (Green Background) - Errors are considered acceptable for most tasks.
  • Average Error: Less than 1.0°
    
  • Maximum Error: Less than 1.5°
    
• FAIR (Gray Background) - Errors are moderate, and the calibration could likely be improved.
  • Average Error: Between 1.0° and 1.5°
    
  • Maximum Error: Between 1.5° and 2.0°
    
• POOR (Red Background) - Errors are generally too high for useful eye tracking in most scenarios.
  • Average Error: Greater than 1.5°
    
  • Maximum Error: Greater than 2.0°
    

The acceptable error threshold depends on your specific research goals and participants.

• For typical adults with healthy vision, we suggest aiming for a stricter standard than "GOOD," such as an average error below 0.5° and a maximum error below 1.0°.
  
• For infants, young children, or impaired populations, a greater margin of error may be necessary and acceptable, depending on the size of your experimental stimuli.