When all models are wrong: More stringent quality criteria are needed for models used at the science-policy interface

In order to ensure quality in the treatment of uncertainty of science for policy more rigorous appraisal methods are needed to substantiate the inference offered via mathematical modelling. Seven rules are suggested to extend the use of classic sensitivity analysis to sensitivity auditing of models used in a policy context. Examples illustrates the various rules. The present articles draws on a more technical paper published on Foresight and Innovation Policy, Special Issue on Plausibility.