Turn an unexpected observation into a bounded investigation with a preserved raw record and a justified next action.
Field diagramA quality-check decision
Estimated reading
8 min
Estimated practice
12 min
Equipment
Notebook, deviation form, instrument and method records, clean water for a controlled demonstration, and approved cleaning materials.
Progress stays in this page session. Allow lesson storage in Cookie settings to remember it; browser settings may also prevent saving.
Objective
By the end, you can:
Describe a deviation factually without guessing its cause.
Use a small, ordered troubleshooting tree.
Decide whether to accept, flag, investigate, or repeat a result while preserving the original.
A deviation is information
A deviation is a departure from the planned method, condition, or record. ‘The pipette made a strange sound’ is a reportable observation; ‘the pipette was broken’ is a conclusion that needs evidence. Write what happened, when, where, who observed it, which sample or control was affected, and what was done immediately. Preserve the original reading and label any later value as a repeat or corrected record.
Troubleshooting should reduce uncertainty one variable at a time. Start with identity and transcription, then equipment status, environment, consumables, technique, and sample or control condition. Stop when a safe boundary is reached. A clean-water demonstration can teach the logic, but it cannot stand in for a method-specific investigation of unknown samples.
Accept, investigate, or repeat
Acceptance means the record meets the predeclared method and quality criteria. Investigation means the result is held while evidence is gathered. Repeat means a new measurement is planned under an identified reason, with the original retained. Repeating until a preferred value appears is selection, not troubleshooting. If the cause remains unknown, report the uncertainty and limitation rather than manufacturing certainty.
Fact: ‘blank was 0.18; threshold was 0.05.’
Hypothesis: ‘carryover may have occurred.’
Action: ‘pause sample interpretation; inspect tip and rerun a blank under the approved rule.’
Quality recap before troubleshooting
Before investigating a suspect result, identify the relevant blank or reference, its predeclared criterion, and the consequence of failure. A blank asks whether the process adds background; a reference asks whether a known value is recovered. If the run has neither, record that limitation and do not invent a pass rule after seeing the result.
Worked example
Preserving a suspect reading
During a water transfer, the operator notices a droplet outside the tip. The next blank is above its predeclared threshold.
01Fact: droplet observed before dispense; blank exceeded criterion.
02Action: mark affected transfer and downstream readings pending; retain raw readings.
03Investigation: inspect tip fit and technique, use a new tip, run the approved blank, and document whether the control returns within criterion.
Materials
Set out what you need.
Deviation form
Instrument log
Approved method or worksheet
Clean water and labeled practice vessels
Safety & stop conditions
Pause if the work no longer fits the plan.
Stop when a spill, damaged equipment, unknown material, or unsafe condition appears.
Do not open, repair, or recalibrate equipment without authorization.
Use only water for practice and follow local waste instructions.
Method
Work through the steps.
01
Freeze the record
Save the raw reading, image or printout if available, sample/control IDs, time, operator, and method version.
02
Describe the deviation
Write observable facts, not a blame statement or unsupported cause. Identify the exact step or condition that differed.
03
Protect affected work
Pause use of suspect equipment or data as the method requires. Mark affected results as pending investigation.
04
Test the simplest causes
Check identity, transcription, units, instrument status, environment, consumables, and technique in that order, changing one variable at a time.
05
Close with a disposition
State accept, investigate, repeat, or invalidate with the criterion and evidence. Link any new run to the original record.
Checkpoints
Observe, record, investigate.
The original value is still visible.
The deviation is time-stamped and tied to an ID.
Troubleshooting tests one plausible cause at a time.
The final decision has a criterion or clearly states that none exists.
Common mistakes
Correct the process, preserve the record.
A bad-looking value is deleted before review.
Preserve it and mark it suspect; deletion removes the evidence needed to assess the problem.
Several settings are changed at once.
Change one condition at a time so the observation can inform the cause.
The same error is repeated with the same setup.
Identify what new evidence the repeat is meant to provide and change the relevant condition.
A deviation form says only ‘operator error.’
Describe the observed action and effect; assign cause only when supported by evidence.
Practice
Troubleshooting tree
Use supplied data: a pipette transfer has an external droplet and the following blank is high. Write the investigation sequence and disposition.
