- Sample: quality, concentration, degradation, contamination, storage, or identity.
- Reagent: lot, age, concentration, buffer, enzyme activity, or compatibility.
- Instrument: calibration, program, temperature, optics, rotor, or plate reader settings.
- Operator: setup order, pipetting, labeling, mixing, timing, or documentation.
Change one thing at a time
When every variable changes at once, a successful repeat does not reveal the cause. A structured repeat changes the most likely factor first while keeping the rest stable.
Use controls as evidence
Controls are not formalities. They tell you whether the system can work, whether contamination exists, and whether the sample is the only weak point.
Practical troubleshooting note
Write the suspected cause, the evidence for it, the change made, and the result. This creates a learning record instead of a cycle of repeated uncertainty.
When to stop optimizing
If a method repeatedly requires extreme conditions, consider whether the design, primer pair, sample, or assay choice is the real problem.