Uncertainty of pesticide residue concentration determined from ordinary and weighted linear regression curve

Statistical evaluation to validate matrix-matched calibration for standardized beany odor compound quantitation in yellow pea flour using HS-SPME-GC-MS

Accurate and precise quantitation of beany odor compounds is important in developing yellow pea (Pisum sativum L., YP) flour-based foods. Aiming at establishing standardized external calibration using an internal standard (ECIS) quantitation method, the effect of solvent extraction on matrix deodorization and systematic statistical analysis on quantitation was evaluated. Initially, accelerated dichloromethane extraction on YP flour and starch produced two clearest deodorized matrix-matched matrices. Secondly, due to the heteroskedasticity, weighted least squares regression (WLSR) was introduced to build calibration curves. The curve linearity and regression parameters were further confirmed via a t-test. Lastly, methodology indicators including LOD/LOQ, accuracy and precision, and the matrix effect (ME) were assessed. Results showed that there were no significant differences in the quantity of beany odor compounds interpolated from two deodorized matrices. This study demonstrated for the first time that despite the unignorable ME, deodorized starch is a feasible and affordable alternative to deodorized YP flour in the quantitation of beany odor compounds to achieve a reliable result.

The inadequate use of the determination coefficient in analytical calibrations: How other parameters can assess the goodness-of-fit more adequately

Simple linear regression using ordinary least-squares is the most common function applied in laboratories for analytical calibrations. The determination and/or the correlation coefficients are usually the parameters applied for assessing the goodness-of-fit of a simple linear calibration. However, these parameters are unable to detect the highly biased results at low calibration levels that are obtained with ordinary least-squares. In this study, the use of other parameters based on the relative standard errors of the calculated contents is evaluated. It has been found that these alternative parameters can detect the biased results obtained at low calibration levels with ordinary least-squares, being the relative standard error the one that seems to provide the most adequate results. Ordinary least-squares should only be applied if the lower limit of quantification is set to at least five times above the conventional limit of quantification. For trace analysis, where the lowest possible limit of quantification is required, weighted least-squares should be applied to obtain accurate estimates, especially at low concentrations. One of the greatest advantages of the relative standard error is that this parameter can be determined for all types of regression functions and is not limited to calibrations with linear relationships between the variables.

Statistical evaluation of analytical curves for quantification of pesticides in bananas

The aim of this paper is to statistically validate the analytical curves of a chromatography method to identify and quantify azoxystrobin, difenoconazole and propiconazole residues in banana pulp, using QuEChERS and GC-SQ/MS. A matrix-matched calibration was used and analytical curves were estimated by weighted least squares regression (WLS), confirming heteroscedasticity for all compounds. Statistical tests were performed to confirm the quality adjustment of the proposed linear model. The correlation coefficient for azoxystrobin, difenoconazole and propiconazole were, respectively, 0.9985, 0.9966 and 0.9997 (concentration range: 0.05 and 2.0 mg kg^-1). The limits of detection and quantification were, respectively, between 0.007 and 0.066 mg kg^-1, and between 0.022 and 0.199 mg kg^-1, below the maximum limits stipulated by Brazilian, American, and European legislation. Only difenoconazole had an insignificant matrix effect (6.8%). Thus, the weighted least squares method is shown to be a safe linear regression model, providing greater reliability of the results.