However, the nonparametric test did not
reject the null hypothesis that the median load of contaminated
chilled carcasses (4.7 MPN/100 g) was not significantly higher the
median load of contaminated frozen carcasses (4.6 MPN/100,
P ¼ 0.056) and those of contaminated freshly slaughtered one
(4.6 MPN/100, P ¼ 0.498), even if 75th percentile of loads of those
contaminated chilled one (21.5 MPN/100 g) was about twice of
those of contaminated frozen carcasses (11.5 MPN/100 g) (Table 3).