Life cycle assessments should be used to test hypotheses rather than to produce absolute values
In a recent open-access article published in the journal PLOSone, researchers at CML, together with colleagues from Vietnam, present strong arguments for why Life Cycle Assessment (LCA) studies should work towards hypotheses and LCA results only ever should be interpreted as relative.
The argumentation is based on that much of the uncertainty underlying results in comparative studies are related to shared processes (e.g. electricity production, refinery, mineral extraction, etc.). These uncertainties are therefore not relevant for the comparative purpose and can be disregarded by adopting dependent sampling. Only the relative difference between the results yielded from dependent sampling can then be compared (A-B) and tested using more powerful paired statistical tests. Using the example of different scales of Pangasius farming in Vietnam, the authors show how the null-hypothesis of A=B could be rejected with a high level of confidence, despite a large overlap in error bars.
The outcomes of the study has implications for comparative LCA studies, product footprints, policy decisions, etc