Interpretation, Uncertainty, Sensitivity and Critical Review
What this domain tests
Interpretation tests whether you can identify significant issues, test robustness, distinguish uncertainty types, draw bounded conclusions, and design a review that matches the study's disclosure level.
Core rules and concepts
- ISO interpretation has three main steps: identify significant issues, evaluate completeness, sensitivity, and consistency, and formulate conclusions, limitations, and recommendations.
- Parameter uncertainty comes from data ranges; scenario uncertainty comes from choices such as grid mix or end-of-life; model uncertainty comes from simplifying assumptions.
- Sensitivity analysis changes one or more key parameters to see whether conclusions move. Contribution analysis shows which processes drive the result.
- Critical review can be internal or external. A public comparative assertion requires a panel review with interested parties.
Worked example
A product carbon footprint is 100 kg CO2e, with a data range of plus or minus 15 kg. Electricity is 60% of the result. A sensitivity case replaces the average grid with a renewable tariff and lowers the result to 70 kg CO2e. The conclusion "electricity is the dominant hot spot" is robust, but the precise headline value is not. The report should present the range, the scenario, and the condition attached to the claim rather than a single false-precision number.
Traps that cost marks
- Reporting only a mean value when key data are uncertain.
- Running sensitivity on trivial parameters while ignoring the dominant process.
- Confusing uncertainty with natural variability or with scenario choice.
- Treating a review as a final formatting check rather than a methodological audit.
- Drawing a comparative conclusion that the sensitivity analysis does not support.