
In order to detect and graphically visualize the absence or presence of systematic errors in fit data, conditional probabilities are employed to analyze the statistical independence or dependence of fit residuals. This concept is completely general and applicable to all scientific fields in which model parameters are fitted to experimental data. The applications presented in this work refer to published charge-density data.
Keywords: charge-density studies; quality indicators; systematic errors; conditional probability; statistical independence.
Supporting information
![]() | Portable Document Format (PDF) file https://doi.org/10.1107/S2053273314012984/eo5030sup1.pdf |