Tests of the reduced width amplitude distribution with proton resonance data

Abstract

The assumption that the distribution of reduced partial-width amplitudes is multivariate Gaussian is tested by separate measurement of width and amplitude correlations. Two approaches to the analysis are employed:1) analyzing the data sets separately and then averaging the results;2) combining the data sets and then performing the analysis. Although the latter procedure provides a nuclear width ensemble (1117 widths), several tests suggest that it is preferable to analyze the individual data sets and then average the results. When the data are transformed to a representation in which the amplitude correlation is zero, the average width correlation for 21 data sets is ¯r w =−0.01+0.03, consistent with the Gaussian assumption.