Algorithm for improving psychophysical threshold estimates by detecting sustained inattention in experiments using...

Area of Science:

• Psychophysics
• Neuroscience
• Experimental Psychology

Background:

• Psychophysical experiments often assume stationary perception, but attention lapses can bias results and affect threshold estimates.
• Existing methods to detect non-stationarities are limited, especially for experiments with fewer than 300 trials.

Purpose of the Study:

• To develop and validate a novel algorithm for detecting inattention on a trial-by-trial basis in psychophysical experiments.
• To improve sensory threshold estimates, particularly when using adaptive procedures like Parameter Estimation by Sequential Testing (PEST).

Main Methods:

• A new algorithm was developed to identify inattention during psychophysical trials.
• The algorithm's performance was assessed through computer simulations and a behavioral experiment with stroke patients assessing proprioceptive difference thresholds.

Main Results:

• Simulations showed up to a 77% reduction in estimation errors for inattentive subjects, even with fewer than 100 trials.
• In behavioral data, inattention was detected in 14% of assessments, and the algorithm reduced test-retest variability in 73% of corrected pairs.

Conclusions:

• The novel algorithm effectively detects inattention in psychophysical experiments, improving threshold estimation accuracy and reducing variability.
• This method can be used post hoc or online, offering practical benefits for experiment duration and data quality, especially in clinical settings.