Specifications

ChemStation Method Validation Pack — Checkpoint Planning
which could influence the result.
Ruggedness of an analytical
method is given if the deviation of
laboratory mean values is not sig-
nificantly different from the devia-
tion of all measured values.
Ruggedness should show the relia-
bility of an analysis with respect
to the influence of external para-
meters. Ruggedness of a method is
distinguished by the fact that
change of parameters within a rea-
sonable range has no significant
influence on the result. The F-test
and the t-test can be applied dur-
ing a statistical test. As a measure
for ruggedness, the comparative
standard deviation is calculated
and listed. Figure 28 shows the
planning dialog of checkpoint
“ruggedness/robustness”.
Planning data
• Determination method:
• Comparison of results
• Comparison with reference
• Other data:
• Number of series ( from 2-50)
• Y-units
• Multiple injection possible
Output settings for calculations
• Neumann trend test
• Dixon or Grubbs test for
outliers
• Variance homogeneity
• Repeatability limit
• Reproducibility limit
• Error of result
• Range of confidence
(repeatability conditions)
• Range of confidence
(reproducibility conditions)
The following results are always
calculated for robustness testing:
• standard statistics such as
mean value, RSD, repeatbility
and reproducibility along with
confidence intervals for
repeatability and reproduci-
bility, and
44
This section lists all checkpoints
with their planning options and a
short explanation of their mean-
ing.
Precision
Precision describes the extent of
conformity between results
obtained during repeated use of a
set analytical method under recur-
rent and comparable conditions.
Monitoring the precision records
random errors. Precision can be
planned as precision in the true
sense, or as repeatability from lin-
earity. In both cases, the nominal
values of the variation coefficient
can be entered. Precision can be
performed with multiple injec-
tions. Figure 27 shows the plan-
ning dialog for checkpoint "preci-
sion”
Planning data
• Determination method:
• Precision
• System precision from
linearity
• Other data
• Number of values ( change
default)
• Y-units
• Multiple injection possible
Output settings
• Repeatability limit
• T-value
• Confidence interval
• Error of result
• Trend test according to
Neumann
• Outlier test according to
Dixon/Grubbs
Robustness/Ruggedness
Robustness is defined as the inde-
pendence of an analytical result
from changes in other parameters,
Figure 27
Planning dialog for checkpoint “Precision”