| Abstrakt: |
Errors in the parameters for total neutron resonances which were based on experimental data for sample transmission have been analyzed by statistical methods. Two fundamental methods for parameter analysis were studied — the least squares method, which is used in the case of good resolution of levels, and the area method (used in the case of poor resolution). For the first method, optimum sample thickness was determined, formulas were derived for calculating parameter errors which take into account not only counting statistics but also the effect of background and other source-contributed errors. For the area method, formulas and tables were worked out which allow a determination of errors in the parameters s0, G, s0 G, s0 G2 for known errors in the total area. A significant loss in precision was evident when two samples were used which differed slightly in thickness. |
Full text from SpringerLink