Analytica Chimica Acta 478 (2003) 281–301 Correction function on biased results due to matrix effects Application to the routine analysis of pesticide residues Luis Cuadros-Rodr´ ıguez a,∗ , Ana M. Garc´ ıa-Campaña a , Eva Almansa-López a , Francisco J. Egea-González b , M. Lourdes Castro Cano b , Antonia Garrido Frenich b , José Luis Mart´ ınez-Vidal b a Department of Analytical Chemistry, Faculty of Sciences, School of Qualimetrics, University of Granada, Avda. Fuentenueva s/n, E-18071 Granada, Spain b Department of Analytical Chemistry, University of Almer´ ıa, E-04120 Almer´ ıa, Spain Received 5 April 2002; received in revised form 11 November 2002; accepted 18 November 2002 Abstract A new strategy to carry out the correction of analytical results affected by systematic errors due to the matrix effect is proposed. Two types of external calibrations must be established with the purpose to estimate the matrix effect: solvent calibration (SC) and matrix-matched calibration (MC). These calibration curves are statistically compared and a correction function (CF) is proposed with the aim to simplify the resolution to the problems associated with the incidence of matrix systematic error in the analytical results. Applying this correction function to the results obtained from the solvent calibration, it is possible to make a prediction of the values that would be obtained when the matrix-matched calibration is applied. On the other hand, a rigorous study of the associated uncertainty is developed and applied to the calculated correction function. Finally, this correction function is validated by means of obtained data of recovery studies carried out by a traditional methodology. The methodology has been satisfactorily applied to the quantification of the pesticide procymidone by HPLC for assessing dermal exposure. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Matrix effect; Calibration; Correction function; Uncertainty; Recovery; Chromatographic analysis 1. Introduction The traditional aim of the calibration in Analytical Chemistry [1] implies the establishment of a func- tional relationship between an instrumental response and a given value corresponding to a standard. Con- sidering that the response can be obtained either from an instrument or from a measurement system, two as- pects of calibration [2] must be taken into account: the ∗ Corresponding author. E-mail address: lcuadros@ugr.es (L. Cuadros-Rodr´ ıguez). first one corresponds to the equipment or instrument calibration; while the second is the process or method calibration, in which the equipment calibration is included. In the case of indirect calibration is usual to estab- lish a calibration function [3] (response: mathematical function of standard values). In general, this last fea- ture is the most widely applied and statistical methods of regression are used in order to find out the cali- bration function (or calibration curve) best fitted to a theoretical mathematical model previously selected [4,5]. The selected calibration curve allows to explain 0003-2670/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII:S0003-2670(02)01508-8