AISM 53, 338-353
© 2001 ISM
(Received December 14, 1998; revised November 8, 1999)
Abstract. This paper presents new omnibus tests for the exponential and the normal distribution which are based on the difference between the integrated distribution function $\Psi(t)=\int_{t}^{\infty} (1-F(x)) dx$ and its empirical counterpart. The procedures turn out to be serious competitors to classical tests for exponentiality and normality.
Key words and phrases: Goodness-of-fit test, integrated distribution function, exponential distribution, normal distribution.