Abstract: | The first cross-impact models were developed ten years ago. Since then, many versions of this technique have been developed. This article proposes criteria for evaluating generic cross-impact models and demonstrates the use of these criteria. It distinguishes carefully between generic cross-impact models and application models. Generic models consist of mathematical definitions and computational procedures; application models consist of a generic model plus appropriately expressed data relevant to the topic being analysed. Four criteria are proposed for evaluating generic cross-impact models: internal consistency, robustness, generality, and clarity. |