Detecting multimodality in saccadic reaction time distributions in gap and overlap tasks

Area of Science:

• Neuroscience
• Cognitive Psychology
• Biostatistics

Background:

• Saccadic reaction time (SRT) distributions frequently deviate from unimodal patterns, exhibiting multiple peaks.
• Understanding these multimodal distributions is crucial for accurately characterizing eye movement control.

Purpose of the Study:

• To develop and validate a statistical approach for identifying and characterizing multiple modes within SRT distributions.
• To determine the number and form of individual peaks in SRT data.

Main Methods:

• A statistical method models the overall SRT density as a sum of basis functions with varying weights.
• The method minimizes the squared difference between measured data and the approximation, accounting for data error.
• The number of basis functions (K) is incrementally determined to fit the data.

Main Results:

• Analysis of six subjects across different saccade tasks revealed five distinct modes in SRT distributions.
• These modes include two 'express modes' (80-140 ms), two 'fast-regular modes' (145-190 ms), and one 'slow-regular mode' (approx. 230 ms).
• The identified modes appeared consistently across different subjects.

Conclusions:

• The statistical method confirms the existence of multiple modes in SRT distributions.
• This approach allows for a more nuanced description of SRT data using characteristic numbers beyond mean and standard deviation.
• The consistent location of modes across subjects suggests underlying, stable neural mechanisms governing saccadic responses.