Importance of skewness and kurtosis statistical tests for outlier detection and elimination in evaluation of geochemical reference materials

摘要: A common characteristic of concentration data compilations for geochemical reference materials (GRM) is a skewed frequency distribution because aberrant analytical data. Rejection outlying results usually required to obtain better estimate mean values. The present work describes the application an approach, based on skewness and kurtosis statistical tests, establish reliable values in study rare-earth element GRM. Frequency histograms initial concentrations these elements differ significantly from normal distribution, as deduced their coefficients. procedure shows that rejection outliers yields distributions which can be used provide with smaller standard deviations large number cases, although many our %RSD are similar (within 5%) literature This has been applied derive new 26 GRM, compared earlier show proposed here provides greater “quality value.” Although most those reported previous literature, some them different (with differences up 40%) few