Avaliação do Milktech na detecção de fraude por adição de água em comparação ao crioscópio

Abstract

This work aims to identify, through statistical analysis using the Kappa methodology, possible correlations between analyzes performed by a new methodology (milktech) and by the standard methodology (cryoscope) in relation to the fraudulent addition of water in raw milk still in the expansion tanks of the properties, before arriving at the dairy, thus avoiding numerous damages such as the financial and the difficulty in discarding the nonconforming product. For this, an apparatus called milktech was used and concomitantly performed in the cryoscope, used today as a standard method for detecting fraud by irregular addition of water in milk. 250 samples were analyzed. Initially the analyzes were performed with the milk received in the dairy, being analyzed by the two methods (milktech and cryoscope) and annotated in a specific form. After this period the samples were analyzed by milktech directly in the expansion tank of the properties of Zona da Mata Mineira, where a sample was collected and taken to the laboratory for analysis in the crosstalk and again by milktech, and the results were also annotated in a specific form. A statistical correlation analysis of the data collected in the field by the SPSS program was performed, indicating a significant correlation of 0.66 (p <0.001) for the data obtained in the dairy industry and of 0.49 (p = 0.007) also significant for data obtained directly from the expansion tank. For both cases an adjustment equation is needed to estimate what the true result would be, since the mean difference between pairs of samples was also significant, being -0.014 (95% CI -0.018 to -0.010) (p <0.001 ) and -0.019 (95% CI -0.027 to -0.011) (p <0.001) respectively.