Tests for randomness of spontaneous quantum decay

Abstract

Analyses of the runs (i.e., uninterrupted sequences of identical outcomes) were made of the temporal succession of decays from three nuclear processes: (1) \ensuremath{\beta} decay of ${}^{137}\mathrm{Cs},$ (2) electron capture decay of ${}^{54}\mathrm{Mn},$ and (3) \ensuremath{\beta} decay of ${}^{214}\mathrm{Bi}$ followed by \ensuremath{\alpha} decay of ${}^{214}\mathrm{Po}.$ Sample half lives vary widely from about 30 yr ${(}^{137}\mathrm{Cs})$ to 20 min ${(}^{214}\mathrm{Bi}).$ For each process, the numbers of runs of length k $(1<~k<~20)$ were determined from three distinct types of generating procedures: (a) runs with respect to a target value, (b) runs with respect to parity, and (c) runs with respect to sequential differences. Comparison of experimental results with those predicted by the theory of runs (on the assumption that every distinguishable arrangement of a temporal sequence of disintegrations is equally likely) provides comprehensive support of the hypothesis that the spontaneous decay of a nucleus is a random occurrence unaffected by past history.