Mann proposed a non-parametric test for trend for independent non-seasonal observations. In this paper we generalize the test to m-dependent seasonal observations. Furthermore, an estimator of trend, based on Kendall's tau, is studied and its asymptotic distribution is proved to be normal. The results are exemplified by a moving average process.
|Journal||Scandinavian Journal of Statistics|
|Publication status||Published - 1988|
Subject classification (UKÄ)
- Probability Theory and Statistics
- non-parametric trend analysis
- seasonal observations
- trend estimator