09-04-2020, 05:02 PM
Currently, the EyeLink Host PC software does not support fully automated drift correction. This is an intentional design choice, intended to protect the quality of your data.
A manual check is required to confirm that any detected "drift" is not simply the participant failing to fixate on the target. Automatically accepting a drift correction would risk including these false positives, which would decrease the spatial accuracy of your data.
Alternative: The "Pseudo Drift Check" Method
If your task involves a large number of short trials, performing a manual drift check after each one can be inefficient. Instead of checking every Nth trial or skipping the check entirely, you can implement a "Pseudo Drift Check." This method involves monitoring the participant's gaze at the beginning of each trial to ensure they have achieved a stable fixation on the target (within a set boundary) for a minimum period.
You can use the EyeLink API to create a "pseudo drift check" within your task. This involves presenting a fixation target and comparing the participant's current gaze position to the target's location, which acts as a gate to verify that the calibration is still accurate before proceeding. If the participant successfully maintains fixation on the target for a set period, the trial continues as normal. However, If the participant fails to maintain fixation within the allotted time, the trial will terminate. Your script can then be programmed to automatically initiate a genuine drift check, or go directly to the Camera Setup screen for recalibration. The task can then be rejoined at the point at which the pseudo drift check failed.
This can be implemented in most stimulus presentation platforms, but the exact syntax and capabilities will vary. The attached template illustrates how to implement a pseudo drift check in Experiment Builder.
Simple_PseudoDriftCheck.ebz (Size: 190.82 KB / Downloads: 57)
A manual check is required to confirm that any detected "drift" is not simply the participant failing to fixate on the target. Automatically accepting a drift correction would risk including these false positives, which would decrease the spatial accuracy of your data.
Alternative: The "Pseudo Drift Check" Method
If your task involves a large number of short trials, performing a manual drift check after each one can be inefficient. Instead of checking every Nth trial or skipping the check entirely, you can implement a "Pseudo Drift Check." This method involves monitoring the participant's gaze at the beginning of each trial to ensure they have achieved a stable fixation on the target (within a set boundary) for a minimum period.
You can use the EyeLink API to create a "pseudo drift check" within your task. This involves presenting a fixation target and comparing the participant's current gaze position to the target's location, which acts as a gate to verify that the calibration is still accurate before proceeding. If the participant successfully maintains fixation on the target for a set period, the trial continues as normal. However, If the participant fails to maintain fixation within the allotted time, the trial will terminate. Your script can then be programmed to automatically initiate a genuine drift check, or go directly to the Camera Setup screen for recalibration. The task can then be rejoined at the point at which the pseudo drift check failed.
This can be implemented in most stimulus presentation platforms, but the exact syntax and capabilities will vary. The attached template illustrates how to implement a pseudo drift check in Experiment Builder.
Simple_PseudoDriftCheck.ebz (Size: 190.82 KB / Downloads: 57)