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

Perihan Yolci Omeroglu, A. Ámbrus, Dilek Boyacioglu

Research output: Contribution to journalArticle


Determination of pesticide residues is based on calibration curves constructed for each batch of analysis. Calibration standard solutions are prepared from a known amount of reference material at different concentration levels covering the concentration range of the analyte in the analysed samples. In the scope of this study, the applicability of both ordinary linear and weighted linear regression (OLR and WLR) for pesticide residue analysis was investigated. We used 782 multipoint calibration curves obtained for 72 different analytical batches with high-pressure liquid chromatography equipped with an ultraviolet detector, and gas chromatography with electron capture, nitrogen phosphorus or mass spectrophotometer detectors. Quality criteria of the linear curves including regression coefficient, standard deviation of relative residuals and deviation of back calculated concentrations were calculated both for WLR and OLR methods. Moreover, the relative uncertainty of the predicted analyte concentration was estimated for both methods. It was concluded that calibration curve based on WLR complies with all the quality criteria set by international guidelines compared to those calculated with OLR. It means that all the data fit well with WLR for pesticide residue analysis. It was estimated that, regardless of the actual concentration range of the calibration, relative uncertainty at the lowest calibrated level ranged between 0.3% and 113.7% for OLR and between 0.2% and 22.1% for WLR. At or above 1/3 of the calibrated range, uncertainty of calibration curve ranged between 0.1% and 16.3% for OLR and 0% and 12.2% for WLR, and therefore, the two methods gave comparable results.

Original languageEnglish
Pages (from-to)1-16
Number of pages16
JournalFood Additives and Contaminants - Part A Chemistry, Analysis, Control, Exposure and Risk Assessment
Publication statusAccepted/In press - Mar 25 2018



  • calibration curve
  • ordinary linear regression
  • Pesticide residues
  • uncertainty
  • weighted linear regression

ASJC Scopus subject areas

  • Food Science
  • Chemistry(all)
  • Toxicology
  • Public Health, Environmental and Occupational Health
  • Health, Toxicology and Mutagenesis

Cite this