ADM design
The so-called ADM design of the working group of German market research institutes is a three-stage random sample .
It is usually the basis of research designs in Germany for regional and national election polls, commercial market and opinion research or social science polls. It is used by GfK , Ipsos , Kantar TNS , Marplan and other institutes. The design was developed by the mathematician Friedrich Wendt (1920–2003). The reason for the development of the ADM design is that it is difficult to draw community samples.
The three-stage random sample with area selection is an area sample. The population consists of all private households in Germany and the registered main residence there .
- First stage: In the first stage, sampling points are selected. The larger the district, the greater the probability of selection ( PPS design = probabilities proportional to size).
- Second stage: Addresses are determined using the random route method ( EPSEM random sample = equal probability selection method).
- Third stage: In the third stage, a household member is selected according to the Sweden key. The larger the household, the less likely it is to be selected.
Unlike the academically oriented Gabler-Häder design , the ADM design is more likely to be used in the commercial sector.
Foreigners and the population in communal accommodation such as hotel guests and institutional population (students, police officers, soldiers, community service providers, members of the order, prisoners, patients in psychiatric clinics, people in facilities for the elderly) as well as Germans abroad are excluded from this procedure . Problems may arise with other groups.
With the ADM design, the interviewer (in contrast to the more complex ALLBUS sample) has a relatively greater influence on the selection of target persons, since there is no precisely defined sample of personal addresses before the field begins. This is to be seen as a disadvantage.
Individual evidence
- ↑ WU-Wien.ac.at: Samples. (No longer available online.) Archived from the original on October 15, 2018 ; accessed on October 15, 2018 . Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice. (PPT, 270 kB)
