ADM-SAMPLING-SYSTEM

In Germany, no directory is generally available that contains the addresses of all private households or individuals, which could be used by market research agencies as a sampling frame. The data collected by the local authorities are only available for surveys considered to be of “public interest”.

The “Arbeitsgemeinschaft ADM-Stichproben Face-to-Face” (a group of agencies that is responsible for the sampling system) closes this gap by providing a sampling frame to member agencies. This frame allows representative face-to-face samples to be drawn for all households in Germany and for all people living in those households. In addition, the main statistical data are provided on a detailed level for this population.

The ADM-Sampling-System (F2F) is designed as an area sample covering all populated areas of Germany. It is based on Germany’s topology, organized by states, counties and communities, the statistical areas within communities described by public data, and the geographical data created for traffic navigation systems. Combining these data, the area sample is made up of about 53000 areas, each containing at least 350 but on average about 700 private households.

Prior to sampling, the areas are first regionally stratified according to counties and so-called BIK types1) resulting in some 1500 strata. Based on this stratification, 128 “nets” are extracted containing 210 areas in former West Germany and 48 in former East Germany. These 258 sampling points (= areas) are drawn proportionally to the distribution of private households. For optimal utilization of the stratified sampling frame, sampling is done using the method for random allocation developed by L. H. Cox.2) The key advantage of this method is that it leads to stratified samples without any accumulation of rounding effects. As one area is drawn for one net only and rounding effects are minimized by the Cox allocation, any selected net may be combined with any other selected net, without issues such as differing selection probabilities or too high rounding differences arising. The ADM-Sampling-System (F2F) provides member agencies with as many nets as they need to carry out their surveys.

In the second step, and where necessary in the third step too, the private households (2nd step) and within them the individuals (3rd step) to be polled, are selected randomly using systematic selection methods with a random start. Such methods are known as “random walk”, “address listing with random selection”, “Kish tableau”, “next/last birthday” and others. (These two steps are performed by the agencies themselves)

Since the sampling is done randomly in all three steps (area sampling, household selection, selection of target persons), this method for face-to-face surveys is based entirely on random

sampling. Therefore surveys based on this ADM-Sampling-System (F2F) fully meet the scientific requirements regarding randomization based on statistical theory.

Hermann Hoffmann,

Spokesperson of ADM Sampling Systems, July 2009

 

1) Types of communities differentiated according to the number of inhabitants and the proximity to the nearest conurbation 2) cf. L. H. Cox; A constructive procedure for unbiased controlled rounding; in: Journal of the American Statistical Association 82, S.520 – 524, 1987

In Germany, there is no generally accessible directory containing the addresses of all private households or individuals, nor is there a directory of all the telephone numbers used by private households. Since no official telephone directory exists listing those members of the population who can be reached by means of telephone surveys, neither complete individual nor statistical data are available for this population.

The “Arbeitsgemeinschaft ADM-Telefonstichproben” (a group of market research agencies responsible for the sampling system) has closed this gap by providing a sampling frame to member agencies. This frame allows telephone samples to be selected representatively with respect to the population living in private households, provided they have a telephone. At the same time, it contains the main statistical data that are necessary to describe this population.

The sampling frame itself – the so-called “ADM-Telefonstichproben-System” – is based on the range of numbers available in the German telephone network as updated, monitored and published on the Internet by the “Bundesnetzagentur” (the government agency in charge of the German telephone network). This range of numbers covers all possible telephone numbers in Germany, whether actually in use or not.

Based on methods developed by Sabine Häder and Siegfried Gabler to generate telephone numbers1), the “Bundesnetzagentur” data are processed as follows:

• Numbers from the German landline-based telephone network are generated as blocks of numbers with a range of 10 (the so-called “Festnetzdatei”)
• Numbers from the German cellular telephone network are generated as number blocks with a range of 10000 (the so-called “Mobilfunkdatei”).

 

Since about 99% of the population can be reached via at least one telephone number, whereby some 12% can only be reached by mobile phone (as at 2011)2), the two sampling frames allow representative telephone surveys of the German population to be conducted at any time. The research project “Dual Frame Approaches”, which was carried out in 2011/12 at the initiative of the ADM, has also established the standards that allow the best possible utilization of the two sampling frames in order to produce representative samples. Using the landline-based sampling frame has a further advantage. Landline phone numbers include regional markers, which allow stratified sampling procedures to be used, differentiating down to the level of the official municipality code (AGS) – or even to the district level in larger cities. The well-known, proven stratification method used in face-to-face-surveys can therefore be used for telephone samples too. Due to the small proportion of listed cellular phone numbers to non-listed numbers, the second sampling frame cannot be stratified according to regional criteria. Instead the dialing prefixes or their allocation to individual providers can be used as external stratification criteria to improve the sampling design.

To best utilize this stratification, sampling is usually done using the method for random allocation developed by L. H. Cox3), whereby the allocation is proportional to the number after the decimal point.

As both the drawing of the telephone sample (and thus the target individual, in the case of cellular phone numbers) and the selection of the target individuals within the private household contacted, in the case of landline samples, are performed randomly, ADM telephone samples can be considered to be representative random samples, meaning that the random models developed in mathematical statistics apply to them.

Hermann Hoffmann

Spokesperson of ADM Sampling Systems

October 2013

1) Cf. S. Gabler, S. Häder; Ein neues Stichprobendesign für telefonische Umfragen in Deutschland; in: S. Gabler, S. Häder, J. H. P. Hoffmeyer-Zlotnik (Ed.); Telefonstichproben in Deutschland, pp 69 – 88, Opladen 1998 2) Cf. Research report on the ADM research project “Dual Frame Approaches” 2011/2012 on the ADM website: https://www.adm-ev.de/forschungsprojekte 3) Cf. L. H. Cox; A constructive procedure for unbiased controlled rounding; in: Journal of the American Statistical Association 82, pp 520 – 524, 1987

 

1. vgl. S. Gabler, S. Häder; Ein neues Stichprobendesign für telefonische Umfragen in Deutschland; in: S. Gabler, S. Häder, J. H. P. Hoffmeyer-Zlotnik (Hrsg.); Telefonstichproben in Deutschland, S. 69 – 88, Opladen 1998
2. vgl. Forschungsbericht zum ADM-Forschungsprojekt “Dual-Frame-Ansätze” 2011/2012; auf der ADM-Website: https://www.adm-ev.de/forschungsprojekte/
3. vgl. L. H. Cox; A constructive procedure for unbiased controlled rounding; in: Journal of the American Statistical Association 82, S.520 – 524, 1987