L4: QA-QC Flashcards
Quality
Features/characteristics of a product/service that meets needs and expectations
Quality Assurance (QA)
system of activities (planning, assessment, quality improvement) to ensure that a process, item, or service reaches quality and expections
Quality Control (QC)
measures attributes and performance of process, item, or service against standards to verify if they meet requirements (protection against out of control conditions and ensures results are of acceptable quality)
Accuracy
measure of closeness of an individual measurement/avg number of measurements to the true value
Assessment
evaluation of performance/effectiveness (audits, performance evaluation, management systems review, peer review, inspection)
Audit (Quality)
examination that determines whether quality activities comply w/ planned arrangements and if they are effectives and suitable
calibration
Comparing measurement standard, instrument, or item w/ a standard or instrument of higher accuracy to eliminate inaccuracies by adjustments
Bias
Distortion of a measurement process that causes errors in one direction (measurement different from sample’s true value)
What are three categories of bias?
constant, proportional, variable
Comparability
measuring one data set to another with confidence
Completeness
comparing valid data from measurement systems to expected data obtained under correct, normal conditions
Confidence Interval
Numerical interval constructed around a point estimate of a population parameter, combined w/ probability statement linking it to the population’s true parameter value
Detection Limit (DL)
measure of method capability to distinguish samples that don’t contain a specific analyte from samples w/ low concentrations of the analyte (analyte and matrix specific)
Method
procedures/techniques to perform an activity (sampling, chemical analysis, quantification)
Performance Evaluation (PE)
Type of audit where data is obtained independently and compared w/ routinely obtained data to evaluate the proficiency of an analyst/laboratory
Precision
mutual agreement among individual measurements of sample property (similar conditions)
Reproducibility
variability among measurements of sample sample at different laboratories
Representativeness
describes how well your sample represents the environmental condition you are trying to measure
Standard Operating Procedure (SOP)
document that details operation method, analysis, or action w/techniques and steps; officially approved as the method for performing routine/tasks
Round Robin Study
different laboratories/analysts analyze the same sample w/same method; results are compared to develop new methods and see how reproducible it is
Peer Review
Critical review of work conducted by qualified individuals to ensure activities are adequate, properly performed and documented, and meet quality requirements. Peer reviews provide evaluation of subjects where methods/measures of success are undefined (research, development)
Data Quality Assessment (DQA)
scientific/statistical evaluation of data to determine if data obtained are of right type, quality, and quantity for intended use
5 steps of DQA process
- Review DQAs & sample design
- conduct preliminary data review
- select statistical test
- verify assumptions of test
- conclusions from data
Data Quality Objectives (DQOs)
statements derived from DQO process. Defines data type and tolerable levels of decision errors
Data Quality Objectives (DQO) Process
planning tool that identifies and defines type, quality, and quantity of data for specified use
Data Quality Indicators (DQIs)
stats and qualitative descriptors that are used to interpret degree of acceptability of data utility to user (Bias, precision, accuracy, comparability, completeness, representativeness)
What are two types of errors?
Systematic (reproducible/happens everytime, bias in process) & Random (non-reproducible, estimable)
Measurement Errors
error due to measurement, calibration, and analysis (can be reduced but never eliminated)
Detection Limit
when method can no longer detect a chemical as concentrations approach zero (more difficult to get accurate measurements)
Can you write zero for instrumental measurement?
No, as for any instrumental measurement we only know that concentration is less than detection limit
Sample Handling Errors
results from sample collection, transportation, and storage; can be minimized w/proper handling procedures
Natural Variability
biggest source of imprecision (cannot be controlled), quantify this variability by taking more samples
Rules to ensure accurate data through calibration:
use appropriate amount of standards (~3 & blank), generate calibration curves w/linear regression (use all pts), avoid data at extremes, know when to include a zero (include if instrument can read zero in blank and if its appropriate for colorimetric procedures), evaluate calibration accuracy (correlation coeff, calibration eq), defined calibration range.
