validation Flashcards
what is method validation
is the process used to confirm if the analytical procedures performs the test that is suitable for its intended use.
what do we need in the method
consistent (repeatability and productivity), reliable (precise) and accuracy (trueness).
what is validation used for
is used to judge reliability, consistency and quality od the analytical result
Why do we need statistics?
- when we did crappy experiment
- and if the measurement are crappy, statistics is needed
what are the type of erroe
- systematic error: group/ bias
- gross error: outlier
- random error : noise
how do you determine propagation error
- addition and substration
- multiplication and division
how to do so is in the studying paper
exercise do question in note page 8 and 9.
how to quantify uncertainty
- top down: this is done by evaluating total method by using references, materials and calibration.
- bottom up: divides methods into smaller steps and determine uncertainty in each step and then combine total uncertainty.
what is quality assurance
is systematic measure necessary to ensure that product will fulfill given demands on quality.
When do we need to validate
- before initial of every routine work
- when transferred to another lab
- when condition are changed outside the scope of original method.
what are the Parameters and tests for method validation
- specificity
- selectivity
- limit of detection
- limit of quantification
-trueness - reliability
-quality - consistency
-stability
-
what is Limit of detection (LOD)
lowest amount of an analyte in the sample that can be detected
Limit of Quantification (LOQ)
lowest amount of an analyte in the sample that can be quantitatively determined with suitable precision and accuracy
what is the Ongoing Routine Tests that must be performed to ensure validation of method
- internal quality control: check if the measurement are repeatable.
- control chart historical: check how history activities
- control chart in laboratory: check the expected value and given value
How to reduce the uncertainty of
determinations
- make signal sample as close to centroid as possible
- make the number for measurement of concentration and number of concentration in calibration curve as many as possible
when to use Weighted Regression
- when heteroscedasticity is expected