Question regarding the blinkreconstruct() algorithm in DataMatrix
For the preprocessing of pupil data, I have been referring to the publication "Methods in cognitive pupillometry: Design, preprocessing, and statistical analysis", particularly the blinkreconstruct() algorithm implemented in DataMatrix.
While examining the algorithm, a few questions arose regarding the rationale behind certain preprocessing steps. Unfortunately, I have not been able to find answers to these questions in the literature or documentation. I would therefore be very grateful if you could help me clarify the following points:
1. Interpolation of invalid data: Why are blinks interpolated, whereas data points identified as invalid based on standard deviation (SD) or the pupil-velocity threshold (gap_vt) are not reconstructed?
2. Order of data exclusion: Is there a particular reason why data points identified as invalid based on gap_vt are not excluded before calculating the mean and standard deviation? Could this be because these values might still contain meaningful physiological information?
3. Determination of the velocity threshold: How was the gap_vt threshold determined? Are pupil-velocity changes exceeding this threshold considered definitively invalid, or does the threshold merely indicate an increased likelihood of invalid data?
I would be very grateful for any insights into the rationale behind these decisions. Thank you for your time and consideration!