Further comments on the (miss-)use of r for testing the linearity of calibration functions
暂无分享,去创建一个
The earlier comments of Dr Huber on the incorrect use of correlation coefficient as a test of linearity of calibration are supported with examples of curved relations leading to apparently acceptable values of r. The discussion of proper statistics that should be used is continued. It is noted that many journals do not allow correlation coefficients to be quoted in defense of linear calibrations.
[1] Walter Huber,et al. On the use of the correlation coefficient r for testing the linearity of calibration functions , 2004 .
[2] D. B. Hibbert,et al. Linearity and the limitations of least squares calibration. , 1997, Journal of chromatography. A.