Informatics Data Processing of Outliers in Environmental Laboratories
Laboratory Informatics
Oral Presentation
Prepared by E. Askew
Askew Scientific Consulting, 2952 155th Street, Muscatine, Iowa, 52761, United States
Contact Information: [email protected]; 563-554-9450
ABSTRACT
The impact of Outliers in the basic calculation for data reliability and validity within an environmental laboratory has been identified as being significant in reference compendiums such as “Standard Methods for the Examination of Water and Wastewater” and ASTM E178-16a. But, large data sets or “Big Data” and the interactions between different analytical methods (i.e. such as the impact of nitrate on TKN analysis) require that a multivariate approach to outlier rejection be pursued. This presentation will focus on varied analytical tests that are found in 40 CFR § 136 and 40 CFR § 141 whose results may interact with each other and what has been addressed in the literature on a multivariate outlier informatics approach to judiciously identify outliers and how to treat outliers in reporting results.
Laboratory Informatics
Oral Presentation
Prepared by E. Askew
Askew Scientific Consulting, 2952 155th Street, Muscatine, Iowa, 52761, United States
Contact Information: [email protected]; 563-554-9450
ABSTRACT
The impact of Outliers in the basic calculation for data reliability and validity within an environmental laboratory has been identified as being significant in reference compendiums such as “Standard Methods for the Examination of Water and Wastewater” and ASTM E178-16a. But, large data sets or “Big Data” and the interactions between different analytical methods (i.e. such as the impact of nitrate on TKN analysis) require that a multivariate approach to outlier rejection be pursued. This presentation will focus on varied analytical tests that are found in 40 CFR § 136 and 40 CFR § 141 whose results may interact with each other and what has been addressed in the literature on a multivariate outlier informatics approach to judiciously identify outliers and how to treat outliers in reporting results.
