Use of Robust Statistics to Describe Plant Lead Levels Arising from Traffic

3 Use of Robust Statistics to Describe Plant Lead Levels Arising from Traffic 

In the field of toxic substances which pollute our environment, special importance must be attached to the heavy metal lead, which, until the late eighties arose primarily from automobile exhaust [EWERS and SCHLIPKOTER, 1984], For quantitative measurement of the impact of this lead, it is essential to establish the correlation between the lead content of the plants and soils, the traffic density, and the distance from the road. 

Quantitative studies by means of parametric statistical methods are, however, often very unreliable because of high environment-related variations very often amounting to several orders of magnitude [FORSTNER and WITTMANN, 1983 EINAX, 1990], In other words environmental data sets often contain values which are extremely high or low, i.e. they are outliers in the statistical sense. Also, because environmental data are often not normally distributed, the application of parametric statistical methods results in distorted reflections of reality. 

Application of robust statistics, especially methods of median statistics, for quantitative description of widely varying values may give information which can often be interpreted better than the results from normal parametric statistical methods. 

The purpose of this study is the use of median statistics for quantitative and interpretable recording of the traffic-related impact of lead on soils [EINAX et al., 1991].  

---