Tables 3 and 4 show that in both streams there are variations
between periods in some fractions, while other fractions remain
constant. To verify that this is not a chance occurrence, a contrast
of hypothesis was considered, as in the generation study. The data
used to carry out the contrast of hypothesis were the generation
rates of each component in g/user/working day. In this case,
because the number of data per sample is very small, normality
cannot be assumed. Consequently, instead of using a parametric
test, such as the t-Student test, a non-parametric test was used,
although it is less robust. Therefore, to detect whether there are
differences in the waste composition in both periods, the Wilcoxon
Rank-Sum Test was used for each component and stream (canteen
and cleaning) with a 95% confidence level (a = 0.05). The contrast
statistical data and their associated p values are shown in Table 5.
Tables 3 and 4 show that in both streams there are variationsbetween periods in some fractions, while other fractions remainconstant. To verify that this is not a chance occurrence, a contrastof hypothesis was considered, as in the generation study. The dataused to carry out the contrast of hypothesis were the generationrates of each component in g/user/working day. In this case,because the number of data per sample is very small, normalitycannot be assumed. Consequently, instead of using a parametrictest, such as the t-Student test, a non-parametric test was used,although it is less robust. Therefore, to detect whether there aredifferences in the waste composition in both periods, the WilcoxonRank-Sum Test was used for each component and stream (canteenand cleaning) with a 95% confidence level (a = 0.05). The contraststatistical data and their associated p values are shown in Table 5.
การแปล กรุณารอสักครู่..
