Neutron time-of-flight experiments have long been used to determine resonance parameters. Those resonance parameters have then been used in calculations of integral quantities such as Maxwellian averages or resonance integrals, and results of those calculations in turn have been used as a criterion for acceptability of the resonance analysis. However, the calculations were inadequate because covariances on the parameter values were not included in the calculations. An effort to correct for that deficiency is documented: The R-matrix analysis code SAMMY has been modified to include integral quantities of importance, directly within the resonance parameter analysis, to determine the best fit to both differential (microscopic) and integral (macroscopic) data simultaneously. This modification was implemented because it is expected to have an impact on the intermediate energy range that is important for criticality safety applications.