Covariance analysis in processing measurements
- 15 Downloads
A new approach is considered to processing interlaboratory test data resulting from several different values for the measured quantity. There is a basic method for the simultaneous estimation of the additive and multiplicative components of the total error, which employs covariance analysis. It is shown that if one of the components is not identified reliably, one can transfer to one of the two models in variance analysis.
Key wordsmeasurements interlaboratory tests covariance analysis variance analysis
Unable to display preview. Download preview PDF.
- 1.G. D. Burdun and B. N. Markov, The Principles of Metrology [in Russian], Izd. Standartov, Moscow (1985).Google Scholar
- 2.Quantitative Description of Uncertainty in Analytical Measurements: Eurochem Guidelines/SITAK, 2nd edition [in Russian], VNIIM, St. Petersburg (2002).Google Scholar
- 3.V. I. Dvorkin, Zavod. Labor., Diagnost. Mater., 65, No. 2, 55 (1999).Google Scholar
- 4.V. I. Dvorkin, Metrology and Quality Support in Chemical Analysis [in Russian], Khimiya, Moscow (2001), p. 197.Google Scholar
- 5.G. Scheffe, Variance Analysis [Russian translation], Nauka, Moscow (1980).Google Scholar
- 6.GOST R ISO 5725-2002, Accuracy (Correctness and Precision) in Measurement Methods and Results} [in Russian]Google Scholar
- 7.V. I. Dvorkin, Zh. Analit. Khim., 48,Issue 7, 1127 (1993).Google Scholar
- 8.V. I. Dvorkin, B. I. Minchenko, and I. M. Timonin, XI Conf. Analytical Atomic Spectroscopy: Abstr., Moscow (1990), p. 154.Google Scholar
- 10.V. I. Dvorkin, Zh. Analit. Khim., 54, No. 8, 790 (1999).Google Scholar
- 11.SAS/STAT User's Guide: Release 6.03, SAS Inst. Inc., NC (1988).Google Scholar