10.10 Zeta Score
179
10.10 Zeta Score
Since both z-score and z
-score do not consider the uncertainty of the results reported
by single laboratory, another extension in denominator leads to zeta score, which
includes the standard uncertainty of the result received from the laboratory. The use
of zeta score is only possible when the laboratory provides the result with an assigned
uncertainty. It is worthwhile to emphasize that the ILC/PT providers often ask for
results together with the uncertainty, which stems from the recommendations of the
ISO/IEC 17043. The evaluation criteria are identical to the previously stated criteria
for the z- and z
-scores.
zeta
x lab − X reference
u 2
x + u
2
X
(10.6)
NOTE! zeta score cannot be used when the reference value is obtained from
the results of the ILC/PT participants.
10.11 E n Score
A less frequently used criterion is E n scores, known also as normalized error. This
score is useful for evaluation the results of a single laboratory versus reference value.
E n X lab −X pt /
√
U
2
lab + U
2
pt
(10.7)
where U lab and U pt are the values of the expanded uncertainty (for k 2) for the
result submitted by the laboratory and for the reference value, respectively.
It is worth noting that in the case of the En number, the expanded uncertainty
values are inserted as denominator, hence this influence the evaluation criteria.
|En| ≤ 1.0 satisfactory performance
|En| > 1.0 unsatisfactory performance.
10.12 Summary
The above discussion on the most important scores used by the ILC/PT providers
indicates that the evaluation of the performance of a single laboratory depends on
the applied criteria. Hence, a very important part of the reports is the inclusion of a
179
10.10 Zeta Score
Since both z-score and z
-score do not consider the uncertainty of the results reported
by single laboratory, another extension in denominator leads to zeta score, which
includes the standard uncertainty of the result received from the laboratory. The use
of zeta score is only possible when the laboratory provides the result with an assigned
uncertainty. It is worthwhile to emphasize that the ILC/PT providers often ask for
results together with the uncertainty, which stems from the recommendations of the
ISO/IEC 17043. The evaluation criteria are identical to the previously stated criteria
for the z- and z
-scores.
zeta
x lab − X reference
u 2
x + u
2
X
(10.6)
NOTE! zeta score cannot be used when the reference value is obtained from
the results of the ILC/PT participants.
10.11 E n Score
A less frequently used criterion is E n scores, known also as normalized error. This
score is useful for evaluation the results of a single laboratory versus reference value.
E n X lab −X pt /
√
U
2
lab + U
2
pt
(10.7)
where U lab and U pt are the values of the expanded uncertainty (for k 2) for the
result submitted by the laboratory and for the reference value, respectively.
It is worth noting that in the case of the En number, the expanded uncertainty
values are inserted as denominator, hence this influence the evaluation criteria.
|En| ≤ 1.0 satisfactory performance
|En| > 1.0 unsatisfactory performance.
10.12 Summary
The above discussion on the most important scores used by the ILC/PT providers
indicates that the evaluation of the performance of a single laboratory depends on
the applied criteria. Hence, a very important part of the reports is the inclusion of a