Illustration of subjects raw performances in the boxplots titled 'Data' (the red lines indicate the median, the blue boxes extend from the upper to the lower quartile values, the whiskers show the most extreme points within 1.5 times the inter-quartile range and the red plus signs indicate outliers) and bootstrapped performances with linear adjustments (graphic lines titled 'Means and linear fits'; vertical lines show the standard deviation of the bootstrapped means and the red lines the linear fits). Distribution plots titled 'Non linear F values' show the distributions of F values measuring the distances between the bootstrapped data and the bootstrapped regression lines. Non central distributions suggest non linear effects; significant effects are marked with a star. For all graphics, numbers below the graphics stand for the different groups: (1) LCD/LLN, (2) LCD/HLN, (3) HCD/HLN, (4) HCD/LLN, (5) Controls. Note that although linear adjustments can be observed based on the ranking of the means, the different behavioural tests have different ranking (in contrast to the behavioural tests illustrated on the figure 6). The last (right end side) graphics show the bootstrapped data (blue circles) with smoothed interpolated data (in green – piecewise cubic hermite interpolating polynomial).
Pernet et al. BMC Neuroscience 2009 10:67 doi:10.1186/1471-2202-10-67