Reproducibility of the Tested Protocols

The reproducibility of the test proeedures is expressed as the coeffieient of variation, or relative range, of the three samples of one type taken in one property  

Ideally, the relative range is 0 (max = min). The value of the relative range m differ between properties and test areas. Therefore the frequency distribution of relative ranges has to be calculated. The ciunulative relative frequency of the relative ranges of the protocols shows the difference in reproducibility of the protocols. The relative range, furthermore, may depend on the lead concentration. 

Example The 90th-percentile of the relative range of PROT is 0.4 in the range of COMP between 5 and 10 pg/1. If the result of PROT is 10 pg/1, we can assiune with 90 % probability that PROT is between 8 and 12 pg/1. 

30MS is comparable to the reproducibility of the reference procedure (COMP). This can be explained by the fact that FF and 3 QMS results are not influenced by consiuner behaviour. 

Example If the median is 0.4, for 50% of all samples the relative range is less 

---