Fig 1. MGS task, subject cohort, and speed-accuracy relationship.

a) Temporal schematic of the memory-guided saccade task employed in this study. b) The age distribution of subjects included in the mixed-longitudinal cohort. Individual sessions are depicted by dots. a-b) Reproduced from Figure 1 (Montez et al. 2017), eLife, published under the Creative Commons Attribution 4.0 International Public License (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/). Return sessions performed by the same subject are connected by solid lines. Male subjects are rendered in gray, females in black. Inset histogram depicts the distribution of return sessions across the sample. c) The U-shaped relationship between reaction time and accuracy in the memory-guided saccade task. The x-axis depicts reaction times, z-scored within a session and task condition. The left-hand y-axis represents saccadic error, also rectified and normalized within session. Empirical measurements are rendered in blue. Error bars represent ±1 standard error of the mean. Behavioral data was adaptively binned so that each data point contains the same number of measurements. The smooth quadratic curve (dashed black) indicates the line of best fit for the non-binned data. The vertical red line indicates global minima of the quadratic curve, the point at which the relationship between reaction time and accuracy changes direction. The right-hand y-axis is associated with the gray curve, which depicts the cumulative distribution of trial reaction times across all sessions.