Determination of sample processing uncertainty in bifenthrin residue analysis in carambola and mango

The uncertainty of bifenthrin residue concentration due to sample processing of carambola and mango were determined by analysing the fruits which had been spiked with the pesticide. The fruits were subsequently homogenised to yield analytical portions which were analysed in 15 g and 150 g masses. The uncertainty of sample processing for carambola for 15 g and 150 g were 15.4% and 5.2% respectively. For mango, the uncertainty of sample processing were 26.6% (15 g) and 9.8% (150 g). The uncertainty values obtained indicated that the sample processing uncertainty was significantly larger as the mass of analytical portion decreased. The derived empirical function for the uncertainty of sample processing for carambola within 15–150 g range is CVSP = (4.1/W)0.5 with sampling constant, KS = 4.1 kg. However, the empirical function of the uncertainty of sample processing for mango could not be derived due to falsification of the null hypothesis in F-test analysis.