Application of Statistical Methods for Analysis of Laboratory Data

摘要

The present study focuses on statistical procedures for assessing the consistency between laboratory measurements of several parameters performed by different methods in different laboratories. The aim is to show that statistics is a powerful tool for analyzing measurement data and has a significant role in engineering practice. The subject of analyses is real data from the mining industry. These include measurements of the concentrations of flotation process products (content of copper in the ore, intermediate products, concentrate and waste) as well as the production of molybdenum (feed, concentrate and waste). In the paper, the statistical analysis data of parallel measurements of the above-mentioned parameters from three independent laboratories involves calculating a linear and rank correlation and analyzing how significant the influence of a laboratory factor is. Moreover, the study includes also one- and two-way analyses of variance of the same data from the same laboratories.