Ruban, V.; López-Sánchez, J.; Pardo, P.; Rauret, G.; Muntau, H.; Quevauviller, P. Fresenius journal of analytical chemistry Vol. 370, num. 2-3, p. 224-228 Data de publicació: 2001-06 Article en revista
A new methodology is proposed for obtaining analytical results free from bias errors, it is bases on a modification of the standard addition method. Bias or systematic errors are taken as the differences between the amounts of analyte in the test portion and the image of the analytical signal on a calibration relationship where the amounts of added analyte act as the predictors. The constant component of the error, the matrix interactive effect and the direct matrix interference are jointly discussed as the main sources of bias. The latter two are considered to be different forms of interference. As an example, the new methodology was applied to the determination of sulphate ion in a magnesium chloride sample (Merck AR) using a spectrophotometric and an atomic absorption spectrophotometric procedure. No direct matrix interference or interactive matrix effect on the analytical results was found.