|
392. JKNIFE: A Nonparametric Statistical Test of Variance
by D. M. Thompson, M. F. Richardson and S. M. Rothstein, Department of Chemistry, Brock University, St. Catharines, Ontario, Canada In those experiments where it is usual to evaluate a result (a model) by comparison of observed data with data calculated from the model, such as determination of structure by lanthanide-induced shifts (LIS) in NMR or by X-ray diffraction, the mathematical requirements for the commonly used Hamilton R-factor ratio test for comparing two calculated models are usually not closely enough satisfied. The nonparametric test in this program compares the distribution of weighted (observed minus calculated) values for two models. Statements which are useful in crystallographic applications are denoted in COMMENT cards. FORTRAN IV (B6700) Lines of Code: 560 Recommended Citation: S. M. Rothstein et al., QCPE 13, 392 (1981). |