Basic HTML Version
34
8.2.3.2.
Data Quality Indicators
Data quality indicators (DQIs) will be defined in the QAPP for each test objective. DQIs are a set of
measurable characteristics
that address the quality of data at the field and lab analytical level, typically
precision, accuracy, representativeness, comparability, and completeness. DQIs have some influence on
determining whether the DQOs have been met, as they help define the level of quality in the data. The
QAPP must define the DQIs appropriate for the BWTS test objectives and the measurement quality
objectives (MQOs; the actual acceptance criteria) placed on the DQIs. The MQOs will be used during
data assessment to determine whether the quality of a data set is acceptable relative to the DQIs. Table 3
illustrates the relationship between DQOs, DQIs, and MQOs. Appendix D provides definitions of DQIs
and the types of QC samples used to measures them.
The test objectives and DQIs will establish the criteria for the selection of field and laboratory procedures,
methods, and equipment. DQIs are defined in MERC QAPPs for reference data as well as for ancillary
measurements that support verification test data when possible. DQIs are typically not required in for
vendor technology data, since the vendor is responsible for specifying the quality measurements to be
made to ensure the integrity of their technology’s outputs.
Considerations include reliability under the intended field conditions or sample matrices, detection limits,
specificity, and sensitivity. For all test activities critical to the achievement of the DQOs, the QAPP will
detail:
•
equipment for each field activity and measurement;
•
analytical methods for each laboratory procedure;
•
sampling and data collection procedures;
•
calibration, operation, and maintenance requirements;
•
QC samples and procedures to be implemented in the field and laboratory and MQOs for each
DQI.
Whenever possible these details should be described in SOPs that are developed by the sampling and
testing team or reference laboratory.
Table 3. The Relationship between DQOs, DQIs, and MQOs
Data Quality Objective
(DQO) →
Data Quality Indicator
(DQI) →
Measurement Quality Objective (MQO)
Qualitative and quantitative
study objectives
•
How ‘good’ does the
study data have to be?
•
How many samples are
needed to determine
Quantitative:
•
Precision
•
Accuracy
•
Sensitivity
Qualitative:
•
Representativeness
•
Comparability
•
Completeness
Project specific acceptance criteria for the
DQIs
E.g.,
•
50 samples are needed to
achieve desired level of
confidence (±30%) that
the attribute is correctly
characterized
E.g.,
•
Precision
•
Accuracy
E.g.,
•
Laboratory duplicates precision < 10%
RPD
•
Blank spike accuracy ±15%