Opening the black box

The paper presents methodological approaches to statistical patent analyses. The references to timescales, countries of origin, and patent offices largely determine the outcome of such analyses; in particular, for country comparisons. For instance, considerable differences appear if results are compared based on priority, application, or grant years. For interpreting the patent figures at specific offices it proves important to consider the geostrategic position of the office. Advanced approaches such as the triad concept lead to more balanced results, but their assessment has to include a consideration of international patent flows. For quality indicators it has to be taken into account that patents are always two-dimensional and have technical and economic aspects. In principle early quality indicators primarily cover the technological content. A further issue is the definition of samples aiming at a large number of suitable documents which is sometimes contradictory to the target of completeness.