Therefore, standard deviation is checked (it must be approximately equal to 0).
As described previously (Feng et al. 2016), standard deviations were checked visually by error bars, and statistical significance was determined by analysis of variance.
Residuals exceeding three standard deviations were checked for data errors.
Any outliers above or below 3 standard deviations were checked against original laboratory reports or chart documents.
Precision was verified by determination of relative standard deviation RSD and accuracy was checked by recovery study.
The deviation from 0.5 ratio was checked by performing a two-tailed Binomial-Test.
The deviation from normal distribution was checked using the normal probability plot of model residuals for each variable.
The average standard deviation of split frequencies was checked for a subsample of runs for each analysis to ensure it was < 0.01; this is the suggested diagnostic for determining when the different runs will produce a good sample from the posterior probability distribution.
The deviation from the normal distribution was checked using the normal probability plot of the model residuals (Q-Q-plot) for each variable.
Besides, the mass transfer between the solid and the liquid was checked and the deviations of the simulation were lower than 8.5%.
