Laserfiche WebLink
i D `^ z F -A <br />The Kendall -tau test for trend, also called the Mann -Kendall test, is a non- <br />parametric, rank correlation method which can be used for evaluating cor- <br />relations between data and time. The test is described in Statistical Meth- <br />ods for Environmental Pollution Monitoring, Richard O. Gilbert, Van <br />Norstrand Company, New York, 1987, and in Nonparametric Statistical <br />Methods, Myles Hollander and Douglas A. Wolfe, John Wiley and Sons, <br />New York, 1973. Its use is recommended in An Evaluation of Trend <br />Detection Techniques for Use in Water Quality Monitoring Programs, J.C. <br />Loftis et al., Environmental Protection Agency, 1989. <br />The Kendall -tau test is nonparametric, meaning that it does not depend on <br />an assumption of a particular underlying distribution. The test uses only <br />the relative magnitudes of the data rather than their actual values; there- <br />fore, missing values are allowed, and VOCs and inorganic constituents <br />which are not detected can still berused in the statistical analysis by <br />assigning values equal to half their detection limits. <br />To determine which of the wells containing VOCs and inorganic con- <br />stituents exhibit statistically significant trends, the Kendall Rank Correla- <br />tion Coefficient Test is applied using the Statview SE computer program <br />(Abacus Concepts, 1988). For this particular monitoring program, the null <br />hypothesis, Ho, is that no trend of VOC or inorganic concentrations exists <br />over time. The alternative hypothesis, HA, is that either an upward or <br />downward trend of VOC or inorganic concentrations exists. <br />pjb\b02\b020104j A - 1 Rev. 0 September 25, 1991 <br />