[ D] The annual innovation survey is conceived as a panel study, meaning that every year the same sample of firms is questioned. To replace companies that disappear from the sample (for example due to closure), a random sample of newly-founded firms as well as firms entering the sample frame (e.g. because of growth or change in economic activity) is taken every two years and used to refresh the sample. The use of a panel makes it possible to observe innovation in firms over a long period. This allows for example to investigate the effects of previous innovation activities on employment, productivity and profitability. The panel thus constitutes a unique source of data, both for scientific analysis and to inform economic and technological policy.
The sample frame consists of all firms with 5 or more employees in mining, manufacturing, energy, construction and a large number of service sectors. The surveyed sample is drawn as a stratified random sample. Stratification criteria are sectors (currently 40), size classes (8 in production sectors, 7 in service sectors) and regions (West and East Germany).
