Chi-Square Goodness-of-Fit Tests Based on Dependent Observations
An attempt has been made in this article to investigate the sampling properties of standard goodness-of-fit chi-square tests based on interdependent observations which are obtained from a strictly stationary and strong-mixing random process. We conclude that the null distributions of the test statistics are largely determined by the multivariate probability structure of the process. A few examples are used to illustrate this point.
Key Wordsgoodness-of-fit tests chi-square tests strictly stationary and strong mixing processes linear processes
Unable to display preview. Download preview PDF.
- Graybill, F.A. (1976). Theory and Applications of the Linear Model. Duxbury Press, North Scituate, Massachusetts.Google Scholar