Figure 7: (Rows 1 and 2) Average trial-by-trial agreement of direction estimates based on report and OKN for two observers.
From: Similar contrast sensitivity functions measured using psychophysics and optokinetic nystagmus
Each graph plots agreement as a function of stimulus contrast (as a multiple of each observer’s threshold) for a single SF. The range of agreement expected by chance is the shaded region, with agreement shown as open symbols, and significantly (p < 0.05) higher and lower levels of agreement plot as filled circles and triangles, respectively. Inset in each figure are the psychometric functions for (solid line) report and (dashed line) OKN; shaded region indicates performance falling below chance. (Row 3) Points of significant agreement and disagreement for all 30 observers; the inset numbers are the total percentage of agreement (top) and disagreement (bottom) estimates that were significant out of the number that could have been significant. Note a clustering of disagreements around lower contrast, high SF conditions. (Row 4) Histograms of number of conditions leading to significant agreement across all observers. Note peaks around 0.75× threshold indicating that this level of uncertainty is optimal for assessing agreement.
