On the use of the factorial moments in the study of discontinuous frequency distributions.
Die Werte \(b_1,\,b_2,\,\ldots,\, b_m\) seien beobachtet. Das \(r\)-te Faktormoment \(\sigma_{(r)}\) sei gegeben durch: \(\sigma_{(r)} = \dfrac1m\displaystyle\sum\limits_{i=1}^m b_i^{(r)}\), wo \(b^{(r)} = b (b - 1) \cdots (b - r + 1)\) bedeutet. Ferner sei \(\delta_r = \dfrac{\sigma_{(r)}}{\sigma_{(r-1)}}\). Bei binomialer Verteilung \(f(x) = \dbinom nx p^xq^{n-x}\), gilt nun \(\delta_r=p(n - r+1)\), also \(\delta_{r+1} - \delta_r= - p\). Bei Poissonscher Verteilung \(f(x) = \dfrac{m^xe^{-m}}{x!}\) gilt \(\delta_r = m\). Bei Pascalscher Verteilung \(f(x) = \dbinom{n + x}n p^{n+1}q^x\) gilt \(\delta_r = (n + r) \dfrac qp\), also \(\delta_{r+1}- \delta_r = \dfrac qp\). Bei hypergeometrischer Verteilung \(f (x) =\dfrac{\dbinom kx \dbinom h{n-x}}{\dbinom {k+h}n}\) gilt \[ \delta_r=\frac{(k - r + 1)(n - r+1)}{(k+h-r+1)}, \] so daß entsprechende Werte von \(r\) und \(\delta_r\) auf einer Hyperbel liegen. Drei Zahlenbeispiele.
This page was built for publication: On the use of the factorial moments in the study of discontinuous frequency distributions.
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2591853)
