Abstract

Monitoring of calibration equation performance is essential if high quality of predicted analytical data is to be sustained. In this paper we outline and illustrate the use of some statistical methods which are well suited for post-prediction data scrutiny. Mean square prediction error is partitioned into three components, viz. mean bias, systematic bias and random error. Reproducibility measures such as concordance correlation (r_c), intraclass correlation (r₂) and correlation between difference and sum (r_{(X – Y)(X + Y)}) are also discussed. Other topics discussed include the maximisation of R², type II regression (both variables with error model) and new graphical displays.

© 1999 NIR Publications

Performance measures of color-difference equations: correlation coefficient versus standardized residual sum of squares

Eric Kirchner and Niels Dekker
J. Opt. Soc. Am. A 28(9) 1841-1848 (2011)

Performance prediction of the AXAF Technology Mirror Assembly using measured mirror surface errors

Paul Glenn, Paul Reid, Albert Slomba, and Leon P. Van Speybroeck
Appl. Opt. 27(8) 1539-1543 (1988)

Performance of minimum-mean-square-error filters for spatially nonoverlapping target and input-scene noise

Bahram Javidi, Guanshen Zhang, Farokh Parchekani, and Philippe Refregier
Appl. Opt. 33(35) 8197-8209 (1994)

Quantitative analysis of bayberry juice acidity based on visible and near-infrared spectroscopy

Yongni Shao, Yong He, and Jingyuan Mao
Appl. Opt. 46(25) 6391-6396 (2007)

Use of pupil-difference moments for predicting optical performance impacts of generalized mid-spatial frequency surface errors

Luke A. DeMars and Thomas J. Suleski
Opt. Express 31(22) 36337-36349 (2023)