ABSTRACT
Methods of Exploratory Data Analysis (EDA) are used for quality control (QC) purposes, data analysis, documentation and mapping. In QA these methods can greatly reduce the number of samples necessary to monitor accuracy and precision. EDA incorporates a great variety of methods that give a better insight into data behaviour prior to the treatment of data by any probabilistic model. QC is generally based on the insertion of a control reference sample and blind duplicates (BD) of project samples at a rate of 1 each in every batch of 20 samples. The aim of QC-procedures is to guarantee reliable and comparable analytical data for a geochemical survey. Data integration and adjustments at later times will only be possible when QC becomes an important part of any geochemical survey. EDA offers numerous graphical methods for data analysis and documentation. BDs are used to calculate analytical precision for any pair of BD-analyses.
