# Repeatability and Reproducibility of a Saccadic Eye Movement Time Test

## TL;DR

This study assesses the reliability of measuring saccadic eye movement times using COI-SV® software, finding moderate interobserver consistency but poor intraobserver repeatability.

## Contribution

The study evaluates the repeatability and reproducibility of saccadic time measurements using COI-SV® software, highlighting the influence of age and sex.

## Key findings

- Saccadic duration increases with age, showing a direct relationship (p < 0.05).
- Interexaminer reproducibility was fair to good, while intraexaminer repeatability was poor.
- Bland–Altman analysis showed agreement limits ranging from −159.0 to 220.3 milliseconds for specific measurements.

## Abstract

Background/Objectives: Reliable and objective assessment of saccadic duration is crucial in sports vision, yet standardized clinical tools remain scarce; therefore, this study evaluated the intraobserver and interobserver repeatability of saccadic time measurements using COI-SV® software, and analyzed the influence of age and sex. Methods: Saccadic duration was assessed in 78 participants using a 20/40 Snellen letter stimulus appearing in four directions (up, down, left, right) at two distances. The shortest response time per direction was recorded. General mean values (total, vertical, horizontal, short, long, and ratios) were calculated. Repeatability was evaluated through a protocol of four test repetitions (two intrasession and two intersession with different examiners). ANCOVA and Pearson correlation assessed sex and age effects. Repeatability indices and Bland–Altman plots were used to determine agreement. Results: Regarding sex, there were no significant differences between men and women. Saccadic duration showed a direct relationship with age (p < 0.05), indicating that older participants had worse saccadic time values (longer times). Overall, intraexaminer repeatability was poor, whereas interexaminer reproducibility was between fair and good. Bland–Altman analysis showed limits of agreement ranging from −159.0 to 220.3 milliseconds (ms) for specific time values and from −87.0 to 122.52 ms for general values, which may be useful in clinical practice. Conclusions: The study shows that the COI-SV® software provides moderate to good interexaminer reliability and poor to acceptable intraexaminer repeatability of saccadic duration measurements, indicating that further refinement and validation are needed before considering clinical implementation.