In this context, a new set of 14 real samples were analyzed in triplicate (n = 3) using the CuO/AgO-based electrochemical
sensor. These were 11 WWTP influents, one WWTP effluent, and two natural waters collected in October 2013 and February 2014. As can be seen in Table 3, the results recorded with both the sensor and the standard method were of the same order of magnitude. Eight of the 14 mean COD values obtained with the sensor were within the confidence interval (95%) provided by the accredited laboratory. Besides, in other three samples, the respective confidence intervals overlapped. The samples with very low COD accredited values, namely, lagoon natural water and the WWTP effluent, were also detected by the sensor. It is also worth mentioning the good precision of the sensor analyses, showing relative standard deviations (RSD) below 10% in all the cases. The accuracy of our sensor using the mean COD values of the three replicates for each sample, except those below the LOD, was compared using a statistical ttest. This is the Student’s t-test for paired samples, which determines if there are systematic differences between two different methods of measurement when the study is made with a set of real samples. This test avoids considering just individual samples, and a t-statistic value of the whole set of differences is obtained. The study revealed no significant differences between
the electrochemical sensor and the standard method. Indeed, the experimental t value of 0.414 was smaller than the tabulated statistical value for 12 samples of 2.201, thus confirming the good analytical sensor performance.
In this context, a new set of 14 real samples were analyzed in triplicate (n = 3) using the CuO/AgO-based electrochemicalsensor. These were 11 WWTP influents, one WWTP effluent, and two natural waters collected in October 2013 and February 2014. As can be seen in Table 3, the results recorded with both the sensor and the standard method were of the same order of magnitude. Eight of the 14 mean COD values obtained with the sensor were within the confidence interval (95%) provided by the accredited laboratory. Besides, in other three samples, the respective confidence intervals overlapped. The samples with very low COD accredited values, namely, lagoon natural water and the WWTP effluent, were also detected by the sensor. It is also worth mentioning the good precision of the sensor analyses, showing relative standard deviations (RSD) below 10% in all the cases. The accuracy of our sensor using the mean COD values of the three replicates for each sample, except those below the LOD, was compared using a statistical ttest. This is the Student’s t-test for paired samples, which determines if there are systematic differences between two different methods of measurement when the study is made with a set of real samples. This test avoids considering just individual samples, and a t-statistic value of the whole set of differences is obtained. The study revealed no significant differences betweenthe electrochemical sensor and the standard method. Indeed, the experimental t value of 0.414 was smaller than the tabulated statistical value for 12 samples of 2.201, thus confirming the good analytical sensor performance.
การแปล กรุณารอสักครู่..
