[en] The adjustment of group cross-sections fitting integral measurements is viewed as a process of uncovering theoretical and/or experimental negligence errors to bring statistical consistency to the integral and differential data so that they can be combined to form an enlarged ensemble, on which an improved estimation of the physical constants can be based. An approach with three steps is suggested, and its formalism of general validity is developed. In step one, the data of negligence error are extracted from the given integral and differential data. The method of extraction is based on the concepts of prior probability and information entropy. It automatically leads to vanishing negligence error as the two sets of data are statistically consistent. The second step is to identify the sources of negligence error and adjust the data by an amount compensating the extracted negligence discrepancy. In the last step the two data sets, already adjusted to mutual consistency, are combined as a single unified ensemble. Standard methods of statistics can then be applied to re-estimate the physical constants. A simple example is shown as a demonstration of the method. 1 figure