Fig. 2
Upper panel: Averages of SP scores (black squares) and their standard errors (black error bars) and the averages of \(1/E_{SF}\) (red circles), normalized to the SP averages of groups 2 (morning) and 3 (afternoon), and their standard errors (red error bars), as a function of the average response times for each clock-time group in Brasilia (Brazilian capital) time (UTC-3). Middle panel: The same for the KSS average scores (black diamonds) and the averages of \(1/E_{SF}\) (red circles) normalized to the KSS averages of groups 2 and 3. Lower panel: Relative Risk of pilot errors in the cockpit for the four clock-time groups [dashed blue histograms with each data point (blue triangles with error bars)] and the fitted averages of KSS (black stars with error bars) and \(1/E_{SF}\) (red pentagons with error bars) as a function of the average response times for each clock-time group. All plots refer to sample 3 (N = 216). Details in the text.
