Questions: Method Validation and Acceptance Criteria

Question 1

A laboratory develops a new HPLC method and, after reviewing all validation data, sets the linearity acceptance criterion at r² ≥ 0.995 — a threshold the data comfortably pass. What is the fundamental problem with this approach?

• A. Nothing — acceptance criteria should reflect actual method performance to be realistic
• B. The threshold is too lenient; pharmaceutical methods always require r² ≥ 0.999 regardless of context
• C. Acceptance criteria must be established before data are collected; defining them after seeing the data invalidates the validation
• D. r² is not an appropriate metric for linearity; the analyst should have used the correlation coefficient r instead

Answer: (C)

Pre-defining acceptance criteria is a scientific necessity, not a bureaucratic formality. Setting criteria after seeing the data — even unconsciously — allows the analyst to choose thresholds the data happen to satisfy, converting a rigorous test into post-hoc rationalization. Option B is wrong because linearity thresholds depend on regulatory context and intended use; there is no universal r² requirement applicable to all methods.

Question 2

A new analytical method achieves spike recoveries of 99–101% across all concentration levels but has a relative standard deviation (RSD) of 9% for replicate measurements. Which validation parameter is failing?

• A. Accuracy — the recovery values are too close together to be meaningful
• B. Selectivity — high variability indicates interference from co-eluting compounds
• C. Precision — the method produces highly variable results despite an accurate mean
• D. Robustness — 9% RSD indicates the method is sensitive to small changes in conditions

Answer: (C)

Accuracy (how close the mean is to the true value) and precision (how reproducible the results are) are independent parameters. A method can be accurate on average — the mean recovery is near 100% — yet imprecise, with individual results scattered widely. High RSD signals a precision failure. Students frequently conflate accuracy and precision; the key distinction is mean vs. variability.

Question 3

A pharmaceutical assay method and an environmental screening method for trace-level pollutants in river water may legitimately have different accuracy acceptance criteria, even when measuring the same compound.

• T. True
• F. False

Answer: True

Acceptance criteria are derived from regulatory context and intended use, not from the analyte itself. ICH Q2 pharmaceutical assay methods typically require 98–102% accuracy because drug potency determinations demand tight control. Environmental screening methods for trace analytes in complex matrices may accept 70–130% recovery because lower concentrations, variable matrices, and different risk tolerances make tighter limits impractical.

Question 4

Robustness testing is performed after the main validation study is complete, as a final sign-off before the method enters routine use.

• T. True
• F. False

Answer: False

Robustness testing is a component of the validation protocol itself, not a post-validation add-on. It deliberately introduces small, controlled variations in method parameters (mobile phase pH, temperature, column lot) during the validation study. If robustness testing reveals that the method fails acceptance criteria under minor perturbations, the method must be revised — which means the validation is not yet complete.

Question 5

Why must acceptance criteria be established before validation data are collected, and what scientific risk arises if they are set afterward?

Model answer: Pre-defined criteria convert validation into an objective, reproducible pass/fail decision anchored to external standards (regulatory guidelines, intended use). Setting criteria after seeing data allows the analyst — consciously or not — to choose thresholds the data happen to satisfy, making the 'validation' circular and scientifically meaningless.

This is the same principle behind clinical trial pre-registration: if you define success criteria after observing outcomes, you can always find a criterion the data meet. In method validation, pre-defined criteria ensure the method is genuinely fit for purpose rather than retrospectively declared so. The criteria should come from regulatory guidelines (ICH Q2, EPA methods, ISO 17025) or from the data quality requirements of the end use.