Fig 1.
The LED light and electronic components were housed in an opaque, black box. On top were mounted a viewing tube and goggles with blacked-out frame.
Fig 2.
Example of CFF-threshold calculation for the method of constant stimuli.
We created and plotted binomials models for each participant. The CFF threshold was then calculated for each participant as the x-coordinate at which the proportion of correct responses was equal to 0.8. Shown here as the intersection point of the two dashed blue lines, where y = 0.8 and x = 57.08.
Fig 3.
Correlation between the different methods.
A: Correlation between MOC and MOL descending, b: Correlation between MOC and MOL ascending, and c: Correlation between MOL ascending and MOL descending.
Fig 4.
Distribution of CFF thresholds for the three different measurements.
The method of constant stimuli and descending method of limits both followed a normal distribution (Shapiro-Wilk normality test: P = 0.12 and P = 0.57 respectively). Due to two influential extremely low values, the ascending method of limits does not follow a normal distribution (P < 0.05).
Table 1.
Comparison of the three CFF measuring methods.
Fig 5.
Boxplots of CFF-thresholds for each method, separated by sex.
T = 1.93 for MOC, t = 0.54 for Asc. MOL and t = 1.96 for Desc. MOL. M = males, F = females.
Table 2.
Correlations between each session for each method.
Table 3.
Overall mean CFF, mean CFF difference between sexes and variance distribution for each method.
Fig 6.
CFF test-retest variability for males and females.
Correlation plots comparing CFF-thresholds between session one and two, session two and three and session one and three for the method of constant stimuli (top row), the ascending method of limits (middle row) and the descending method of limits (bottom row). Male CFF displayed in blue, female CFF displayed in purple.
Table 4.
Model summaries of multiple-measures CFF per method for each sex.