期刊
AQUATIC LIVING RESOURCES
卷 22, 期 2, 页码 173-185出版社
EDP SCIENCES S A
DOI: 10.1051/alr/2009019
关键词
Quantiles; Runs test; Mann-Kendall statistic; Spearman correlation; Jonckeere test; Nonparametric slope estimator; Cochran Q; Aligned rank test; Dietz-Killeen multivariate trend test
资金
- EC
- UK government's Department of Environment, Food and Rural Affairs
A selection of ten uni- and multivariate, nonparametric statistical methods thought to be useful for assessing trends in indicators estimated from trawl surveys is described. Nonparametric methods make minimal assumptions about the data and can therefore be suitable when parametric modelling methods, as typically preferred by fisheries scientists, are not suitable. The various methods described are sensitive to a variety of specific features of a trend thereby allowing many different forms of variation over time to be identified, e. g. changes in level, sloping straight or curved lines, and differing cyclicities. In contrast, model-based investigations can be constrained in this regard if the single model which is adopted and fitted to the trend is a simplification of its form.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据