Statistisches Bundesamt (Destatis)

F35 Mikrozensus – Methodik und Aufbereitung SILC und IKT

Statistisches Bundesamt (Destatis)

Zweigstelle Bonn

Graurheindorfer Straße 198

53117 Bonn

The EU survey on the use of ICT in households and by individuals is an annual survey conducted since 2002. In Germany, it has been conducted since 2002.

In 2022, the survey collects data on the access to information and communication technologies (ICT), on the use of the internet, e-government, e-commerce, internet of things, as well as green ICT.

National language: Mikrozensus 2022 - Kernprogramm und Erhebungsteil zur Internetnutzung 

English:

Questionnaire(s) in national language(s) and the translation in English are available in the annex.

The following common concepts and definitions apply under the Integrated European Social Statistics (IESS):

• the International Standard Classification of Education (ISCED) 2011 published in the following breakdowns: low (ISCED levels 0-2: no formal education, primary education or lower secondary education), medium (ISCED levels 3-4: upper secondary or post-secondary non-tertiary education) and high (ISCED levels 5-6: tertiary programmes which normally need a successful completion of ISCED 3 or 4, or second-stage tertiary education leading to an advanced research qualification);
• the International Standard Classification for Occupation ISCO-08 at the 2-digit level;
• the Classification of Economic Activities (NACE Rev.2-2008), at section level;
• the Common classification of territorial units for statistics (NUTS 1) – finer granularity of NUTS 2 is provided on optional basis by some Member states;
• the SCL - Geographical code list;
• information about household income is provided at lower level of detail. 

Additional classifications used in the national questionnaire: None

The ICT survey in households and by individuals covers those households having at least one member in the age group 16 to 74 years old. Internet access of households refers to the percentage of households that have an internet access, so that anyone in the household could use the internet.

No differences in scope at national level.

The survey is collecting data of internet users, individuals who have used the internet in the three months prior to the survey. Regular internet users are individuals who used the internet, on average, at least once a week in the three months prior to the survey.

This annual survey is used to benchmark ICT-driven developments, both by following developments for core variables over time and by looking in greater depth at other aspects at a specific point in time. While the survey initially concentrated on access and connectivity issues, its scope has subsequently been extended to cover a variety of subjects (for example, the use of e-government and e-commerce) and socio-economic analysis (such as regional diversity, gender specificity, differences in age, education and the employment situation). The scope of the survey with respect to different technologies is also adapted so as to cover new product groups and means of delivering communication technologies to end-users.

For more details on the methodology applicable in each survey year, please consult the Methodological Manual for the respective year on CIRCABC - Methodological Manual - Information society statistics (europa.eu).

Deviations from standard ICT concepts: None.

Households and Individuals

In the ICT usage survey 2022, the target population for the different statistical units is:

- individuals: all individuals aged 16 to 74;

- households: all (private) households with at least one member aged 16 to 74. 

Target population composed of households and/or individuals:

• Number of households:  35 343 659
• Number of individuals:  61 257 268

Non-compulsory age groups also included in the target population:

Households:
The total number of households of Germany is about 40.6 million households.
The Non-target groups consist of about 5.3 million households, they comprise all households ...
- without members aged between 16 years and 74 years,
- not living at main residence.
The resulting percentage of under-coverage (non-covered number of households compared to the total country) is about 13 %.

Persons:
The total population of Germany is about 83.5 million persons.
The Non-target population consists of about 22.3 million persons, it comprises all persons ...
- younger than 16 years or older than 74 years,
- not living at main residence,
- living in institutions.
The resulting percentage of under-coverage (non-covered population compared to the total country) is about 27 %.

The data refer to the whole territory of Germany.

All 16 Federal States are covered: Schleswig-Holstein, Hamburg, Niedersachsen, Bremen, Nordrhein-Westfalen, Hessen, Rheinland-Pfalz, Baden-Württemberg, Bayern, Saarland, Berlin, Brandenburg, Mecklenburg-Vorpommern, Sachsen, Sachsen-Anhalt, Thüringen.

Year 2022

Not applicable

Percentages of ‘Households’ and Percentages of ‘Individuals’

Last three months before the interview.

1 March 2022 - 17 June 2022

The legal basis for the 2022 EU survey on the use of ICT in households and by individuals is the Regulation (EU) 2019/1700 of the European Parliament and of the Council of 10 October 2019 establishing a common framework for European statistics relating to persons and households, based on data at individual level collected from samples (OJ  L 261 I, 14.10.2019, p. 1), as implemented by the Commission Implementing Regulation (EU) 2021/1223 of 27 July 2021 specifying the technical items of the data set, establishing the technical formats for transmission of information and specifying the detailed arrangements and content of the quality reports on the organisation of a sample survey in the use of information and communication technologies domain for reference year 2022 pursuant to Regulation (EU) 2019/1700 of the European Parliament and of the Council (OJ L 2269, 27.07.2021, pp. 1-45).

Complementary national legislation constituting the legal basis for the survey on the use of ICT in households and by individuals: 

German microcensus law of 7 December 2016: The up-to-date wording of the relevant national legal act can be found at https://www.gesetze-im-internet.de/mzg/index.html.

The Federal Statistics Law (BStatG) regulates the cooperation and the exchange of data between the Federal Statistical Office and the State Statistical Offices.

See here for the German language BStatG version. 

The up-to-date wording of the relevant national legal provisions can be found at https://www.gesetze-im-internet.de/ (search for terms "Bundesstatistikgesetz" or "BStatG") 

The individual data collected are always kept confidential in accordance with Section 16 of the Federal Statistics Act (BStatG).

Individual data may be passed on only in exceptional cases explicitly regulated by law. Individual data may always be transmitted to:
– public agencies and institutions within the official statistics network which are entrusted with the production of federal or European statistics (e.g. the statistical offices of the Länder, the Deutsche Bundesbank, the Statistical Office of the European Union [Eurostat]),
– service providers with whom a contractual relationship exists (e.g. Federal Information Technology Centre (ITZBund), computer centres of the Länder).

Pursuant to Section 16 (6) of the Federal Statistics Act, institutions of higher education or other institutions tasked with independent scientific research may, for the purpose of carrying out scientific projects, be provided
1. with individual data if attributing the anonymised individual data to the relevant respondents or persons concerned requires unreasonable effort in terms of time, cost and manpower (de facto anonymised individual data),
2. with access to individual data not including name and address (formally anonymised individual data) within specially protected areas of the Federal Statistical Office and the statistical offices of the Länder, if effective measures are in place to safeguard confidentiality.

Persons receiving individual data are also obliged to maintain confidentiality.

The first names and surnames of the household members, the contact details of the household members, residential address, location of the dwelling in the building, first name and surname of the main tenant/owner-occupier of the dwelling, name and address of the household members‘ places of work, and the building age group are auxiliary variables which will only be used for the technical conduct of the survey. As soon as the survey and auxiliary variables have been checked for conclusiveness and completeness, the auxiliary variables will be separated from the information on the survey variables and will be kept separately or stored separately
– Pursuant to Section 14 (5), first sentence, of the Microcensus Act, the first names and surnames and the municipality, street, house number and contact details of the persons surveyed may also be used with regard to household relationships to conduct follow-up surveys in accordance with Section 5 (1) of the Microcensus Act.
– Pursuant to Section 14 (5), second sentence, of the Microcensus Act, the information on the variables pursuant to Section 14 (5), first sentence, of the Microcensus Act may also be used as a basis for recruiting suitable persons and households to conduct household budget surveys and other voluntary surveys.

Information on the survey variables is processed and stored for as long as necessary to comply with the legal obligations. All survey documents as well as the auxiliary variables and the reference numbers originally allocated will be destroyed or deleted after the processing of the last follow-up survey has been finished.

Dissemination of results (publication rules):
Annual results based on a value of less than 71 observations in the sample are replaced by a slash ("/") in published microcensus tables due to the size of their relative standard error (on average over 15%) and the associated low informative value. With 71 to 120 sample observations, the results are shown in brackets because of the greater uncertainty (relative standard error between 10% and 15%).

An EU micro data file (EU scientific use file Germany) will be made available by Eurostat, after approval and permission provided by the German FSO. 

The German microdata files suppress NUTS 1 data and aggregate household size classes with 6 or more household members into one size class 6+. 

There is no special release calendar for the statistical outputs of this survey, but there is a publicly available short-term release calendar for all of the FSO's press releases being released in the following week.

All statistical products from this survey can be derived from the official website of the German FSO.

A weekly preview of all planned press releases of the FSO for the next calendar week is available on the FSO's website.

As soon as the results are final, they are released by the German FSO via different channels, e.g. press releases, internet content, series/tables.

The FSO's policy provides for the dissemination of statistical data to all users at the same time.

As concerns press releases, the FSO informs the Federal Statistical State Offices about the topic and the content some days in advance.

As concerns the public, the FSO presents a short-term release calendar, i.e. a weekly preview of all planned press releases for the next calendar week on the FSO's website.

Annual

Press releases, internet tables and/or other statistical products of this survey are regularly published as soon as final results are available. 

A press release with the first results of the 2022 survey was released on 14 December 2022.

Destatis series "Fachserie 15 Reihe 4" on the 2022 survey (in German language only) was released on 14 December 2022.

Please note: The Destatis series "Fachserie 15 Reihe 4" will be last offered by Destatis for the 2022 survey. From 2023, this series is no longer produced by Destatis.

Standard tables with all Federal results of the survey (i.e. results for Germany as a whole) are sent to customers by the FSO on request from 15 December 2022 on. 

The most relevant statistical outputs are entered into the national data base Genesis Online

(Select theme 1 Territory, population, labour market, elections/12 Population/122 Microcensus/12231 ICT usage in households to go directly to the ICT indicators)

Information not available.

An EU micro data file (EU scientific use file Germany) will be made available by Eurostat, after approval and permission of the German FSO. 

From time to time, the results are also disseminated via press conferences covering various topics, among them ICT-related topics. 

Some cross-sectional publications of the German FSO include results and documentation on the ICT survey, e.g.

- Data Report (in German language)
- Education Report (in German language)
- Report on the German UN SDG Indicators

Information not available.

Scientific article "Die Neuregelung des Mikrozensus ab 2020" on the new microcensus system (in German language with an abstract in English language).
December 2019 issue, released on 16 December 2019.

Scientific article "Die Hochrechnung im Mikrozensus ab 2020" on methodological issues (e.g. grossing up) in the German microcensus from 2020 on (in German language with an abstract in English language).
December 2021 issue, released on 14 December 2021. 

Besides various IT specific manuals, there are mainly the following handbooks for the Federal States and for the interviewers (all in German language; not publicly accessible via internet; attached here as Annex files):

- Introduction to the microcensus survey
- Preparation of the microcensus operation
- Instructions for visiting the selected microcensus districts
- Instructions and Explanations on the questionnaire(s)

Information not available.

The national quality report of this survey (in German language only) is available on the FSO's website here. (Tick theme "Qualitätsberichte" on the left grey column.)

Procedures to promote general quality management principles in the FSO:

The quality manual of the Federal and State Statistical Offices describes the framework for ensuring data quality in German official statistics. It informs users of statistical data (e.g. from the ministries, associations, science or the public) about management to ensure the quality of statistical results. In addition, it serves the employees of the Statistical Offices of the Federal and State governments as well as other bodies in Germany, which compile official statistics, as a guide.

The quality manual (German language) and more information on the FSO's quality guidelines are available on the FSO's website.

Quality assurance framework applied for the survey:

There are several training courses for interviewers as for internal staff responsible for the operation, both in the FSO and in the Federal Statistical State Offices.
The entire production process of the German official microcensus including the different subsamples like the ICT survey is constantly monitored and subject to close-knit business controlling in the FSO.
Meeting EU deadlines for data delivery and for reporting are also subject to the FSO's internal controlling. Developing optimization measures are a constant part of the ongoing work. 
Meetings with national data users on core aspects of the German microcensus such as data quality and user's needs are to follow soon in the next years. 

Planned improvements:

- Further reduction of burden: The FSO is working on further streamlining the entire microcensus questionnaire. For that purpose, it has set up an internal working group specifically responsible for the further development of the microcensus. This body of experts deals, among other things, with planned legal measures/changes and methodological issues. It constantly works on optimizing the microcensus.
- Electronic data collection modes: The quality/validity and the user-friendliness of the questionnaires, which are offered to the respondents in electronic form, are constantly being further developed. 

The new integrated German microcensus entered into force in 2020, whereas the ICT subsample joined that system in 2021. The whole microcensus system is still constantly further developed and assessed. So, solid and stable overall assessments would require some more survey years to come.

As concerns the ICT subsample 2022, the overall statistical output data quality is seen as very satisfactory. The net sample size that could be achieved is sufficient and serves well for fulfilling the EU precision requirements. The indicators produced from the ICT survey 2022 seem very plausible to the FSO, as concerns the whole territory of Germany. One weakness of the survey is that the total sample size does not allow more detailed breakdown results for small Federal States (e.g. Bremen, Saarland, Rheinland-Pfalz): the case numbers in detailed breakdowns are usually too low to enable those Federal States to publish much of the survey's output. For German and EU policy data users, the output of the survey is highly important. The data were finalized right in time and were delivered to the EU in September 2022. The first results were published on 14 December 2022 (press release and internet tables). The series with all results will be released by the end of 2022. The results are comparable across the EU. Compared to previous survey years, the German results of the ICT survey 2022 are comparable with the 2021 survey's results. They are of limited comparability with earlier survey years up to 2020, due to the fundamental system change that took effect from 2021.   

One of the main purposes of the microcensus is to fill the data gap between two censuses. A series of smaller surveys of empirical social and opinion research as well as official statistics uses the microcensus as an extrapolation, adjustment and control instrument. The results of the microcensus - and here especially those of the ICT subsample - provide essential input for the German social policy and digitalization policy. 

Main data users: Parliament, ministries, scientific institutions, social partners, the European Commission (e.g. DG CNCT, JUST, EAC, JRC), Market and Opinion research and the media.

There are no user satisfaction surveys, but regular consultations and information exchange with relevant Federal Ministries and other bodies of experts of the German governmental administration.

User interests are taken into account in many different ways. The federal and state ministries can directly influence the survey program via the legislative procedure for the microcensus.

Furthermore, the data requirement, for example from science or city statisticians, can be found in the Statistical Advisory Board, at microcensus user conferences and specialist committee meetings.

Mandatory variables (100%): All of the mandatory variables required for transmission have been included in the microdata.

Optional variables (close to 100%): As concerns the individuals data set, all of the optional variables have been included in the microdata. As concerns the household data set, most of the optional variables have been included in the microdata. The only exemptions are the optional sociodemographic breakdowns NUTS2 and NUTS3 which were not provided.

Close to 100%.

Since participation in the survey is voluntary, non-response is a source for errors.

In comparison to the obligatory Microcensus population, slight underreporting can be seen in the expected variable characteristics (non-German citizenship, low and high education, some age groups).

Cluster sampling as it is done in Germany raises the error. On the other hand, the stratification reduces it. 

The sampling error reflects the fact that only a particular sample was surveyed rather than the entire population. It is estimated by the standard error and can be expressed by the square root of the estimate of the sampling variance. The estimation of the sampling variance should ideally take into account the sampling design (e.g. the stratification).

More information on methodology for calculating precision estimates is explained below.

Estimation method(s) for the random variation of an estimator due to sampling: Taylor linearization

Tools used to estimate sampling errors: ETOS (further development of CLAN)

Sampling error – calculation methods for the variance estimation (SE): The variance of the weighted total of the variable Y is estimated by the standard variance formula of stratified sampling. But instead of using the value yk of the variable of interest Y for unit k, the weighted residuals zk are considered.

Methods used to assess the standard errors take into account the following specific effects: 

• Implicit stratification: Stratified sample. Error estimation sums over all strata.

• Calibration: Using variance formula on weighted residuals.

References:

• Literature on error estimation:

Andersson, C. (2012). ETOS 2.0 User’s guide. Statistics Sweden.

Qualitätsbericht Mikrozensus 2018 (see Anhang B).

• Literature on sampling:

Bihler, Wolf/Zimmermann, Daniel. Die neue Mikrozensusstichprobe ab 2016. In: WISTA Wirtschaft und Statistik. Issue 6/2016, page 20 ff.

Precision estimates for the question "Individuals having ordered goods or services for private use over the internet in the last 12 months"
(individuals who ticked 'Within the last 3 months' or 'Between 3 months and a year ago' in question D1 of the 2022 model questionnaire):

Number of respondents (absolute value for ‘Yes’ answers):   12 160

Estimated proportion (in %):   76.2

Standard error (in percentage points):   0.37

Details of the breakdowns are available in the Annex below. 

See more details on non-sampling error below.

A register of all dwellings surveyed during the 2011 census formed the sampling frame for the new microcensus sample from 2016. The sample is updated each year by the sampling of new building permits. A small under-coverage of homeless or people not registered and encountered at their dwelling can be assumed but is hard to quantify. On the other hand, people with several residences have a higher probability to be sampled and are therefore over-covered. Likewise the extent has not been quantified. (For the ICT-survey, persons registered as living at their second residence are excluded, due to the defined EU scope).

As the microcensus is a decentralised statistics in Germany, the Federal Statistical Offices of the regions (Federal Statistical State Offices, i.e. the statistical offices of the German "Länder") are conducting the survey. Oftentimes, the survey population at an address is crosschecked with the population register in order to minimize under-coverage.

Not applicable. Sampling is done at the level of areas, not at the level of households/persons.

Not requested in the ICT survey.

1)       Measurement errors:  

Measurement errors can result at all levels of the statistics production. Main sources of such systematic errors that may have occurred in this survey are: 

• deficits of the questionnaire design or instructions given to respondents;
• interviewers' weaknesses in understanding or in conducting the interview;
• incorrect information provided by respondents and/or by interviewers;
• data entry errors;
• errors in IT programmes (data entry, data processing).

2)       Questionnaire design and testing:  

• Designing the questionnaire is a long-lasting process within the FSO, including the Federal Statistical State Offices, ministries and national bodies of expert. In concrete terms, it covers mainly:

• Translation of the English language model questionnaire and adaption to a national language questionnaire (including elaboration of and search for national examples and instructions for respondents);
• Fitting the national ICT questionnaire into the broader microcensus system questionnaire, differentiating between mandatory microcensus core parts (covering e.g. the sociodemographic questions) and voluntary ICT subsample parts;
• Consultations with relevant ministries and bodies of expert on special content and implementation issues;
• Designing/adapting the national questionnaire to different collection modes (CAPI, CATI, CAWI, PAPI);
• Translating the national questionnaire into an English language national questionnaire (for foreign language respondents);
• Establishing the documentation parts on legal bases and legal review of the national questionnaire (FSO's legal department);
• Consulting the Federal Statistical State Offices in order to coordinate and adopt the final national questionnaire;
• Designing the paper version (PAPI) of the finalized national questionnaire with a special,

FSO-wide standardized software application (InDesign).

• Testing of the electronic mode questionnaire(s) by the FSO and by the Federal Statistical State Offices:

Errors arising from the multi-mode data collection are reduced as much as possible by intense tests before the application is released. In addition, efficient filtering controls and plausibility checkings are applied in the electronic modes.

3)       Interviewer training:  

After the COVID-19 pandemic, less CAPI interviews than before the pandemic were personally undertaken, but rather done via telephone (CATI) instead.

During preparation of the survey, the interviewers received intensive training, carried out by the Federal Statistical State Offices. The Federal Statistical State Offices themselves had also received special training and explanatory documentation by the FSO. In addition, various documentation (manuals) were elaborated by the FSO and made available for the Federal Statistical State Offices and the interviewers recruited by them. 
These documents are attached as Annexes under sub-item 10.6.

4)       Proxy interview rates:  

Number of Proxy Interviews: 3 157 

Number of all interviews: 16 444

Share Proxy/Total: 19.2%

Information about non-respondents: 

Main characteristics of non-respondents: Non-response (eligible cases) occurs to a somewhat greater extent when the following characteristics are given: 

• increasing household size,
• increasing number of children in household, 
• decreasing educational status,
• young and middle-aged persons (<50 years old),
• students/pupils, housewives/housemen, persons fulfilling domestic tasks.

The unit response rate is the ratio of the number of in-scope respondents (= the number of achieved interviews or the net sample size to the number of eligible elements selected from the sampling frame).

Unit non-response rate for

• Households:   35.7%
  
• Individuals (aged 16-74):   42.6%

 Comments, if any:

* Sampling is based on areas, not on households. Hence the gross sample is an estimate.

1)       Methods used for dealing with unit non-response: No imputations were made for unit non-response in this survey.

2)       Methods used for minimizing unit non-response:  The methods used for minimizing unit non-response are mainly:

• Press releases and comprehensive information on the Federal Statistical Offices' websites, prior to start of the survey;
• Selected respondents receive advance notification in the form of a letter;
• Dunning procedure for mandatory parts of the survey (i.e. sociodemographic information and some of the questions belonging to the ICT survey);
• IT-controlled system of reminders (organized upon reference dates/ deadlines);
• Phone-calls, postal reminders;
• Bonus payments for voluntary parts of the survey;
• Brochures and other material showing the respondents how the data are used;
• Intensive training of interviewers.

3)       Substitution permitted:  Yes. 

4)       Substitution rate (in %): 19.2% proxy interviews. The criterion used is to select either another household member or a friend or relative. 

Items with low response rates (observed rates in %):  Non-response does only concern the voluntary part of this survey.

In particular, all sociodemographic content and some ICT-related questions of the ICT questionnaire are covered by the mandatory core programme of the German microcensus. If that mandatory information is not provided by the respondent, the dunning procedure starts automatically. Where item non-response in mandatory variables occurs in the final and processed data (e.g. ISCED level of a person), that results from the fact that the respondent in question is still in the dunning process. Usually, none or only very few such cases occur in the processed data.

Items with low response rates (>= 10%) and their observed rates (in %):

• Socio-demographic background EU target variables: None.
• Household-based ICT related EU target variables: None.
• Individuals' based ICT related EU target variables:
• TIME: Interview duration (sample: 14.0%; weighted: 12.4%)

There were no significant processing errors in this survey, but only of minor significance errors which did not impact the final results. Those errors (e.g. data entry errors) could usually be traced back/verified, due to the very complex IT applications behind.

The whole IT process (data entry, data treatment, final data processing) was accompanied by comprehensive and exhaustive plausibility checks and consistency checks. 

Not requested for ICT Survey

The first results (press release, internet tables, series) were released on 14 December 2022.

Along with that, from this date on all results are sent to customers on request. Online data base content on the 2022 survey will be made available in the FSO's 'Genesis Online' data base. 

Time lags (in days) between the actual delivery of the data and the target date when it should have been delivered:

21 days.

21 days (100%).

The statistics is comparable for the whole territory of Germany, i.e. it is comparable between the 16 Federal States of Germany.

The statistics is also EU-wide comparable.

Not relevant

Possible limitations in the use of data for comparisons over time: 

2022 versus 2021: There have been no changes in which case this should be reported.

2022/2021 versus survey years up to incl. 2020:
Please note: According to the German microcensus law, the integration of the ICT households and individuals survey into the German microcensus (1% random sample of the German population) took effect in 2021. Before, the survey was designed as a quota sample with only paper-based data collection mode. The new design is a random sampling-based design, which is closer to reality as concerns the dynamics of the digital society in Germany. Moreover, it offers respondents to use different data collection modes. Due to the fundamental system change, the results of the survey years 2022 and 2021 cannot be directly compared to those of the previous survey years up to 2020.

The length of comparable time series depends on the module and variable considered within each of the modules of the survey.

Not applicable

Not applicable

Not applicable

All statistics are coherent within the dataset.

MANDATORY questions in the Eurostat model questionnaire 2022:

The table in the annex lists the questions that do not reflect the coverage of subjects and characteristics of Annex 2 of the Commission Delegated Regulation (EU) 2021/1898 of the 20 July 2021. 

Adoption of OPTIONAL questions and items in the Eurostat model questionnaire 2022:

The table in the annex lists the optional questions from the annual Eurostat model questionnaire 2022 included in the national questionnaire and their coverage for age groups beyond the standard scope. 

Additional questions introduced in the national questionnaire: 

No additional ICT-related questions have been added to Eurostat MODEL questionnaire.

Effects of deviations from the routing used in the Eurostat model questionnaire:  

All statistics are coherent within the dataset.

No data revision for this statistics in 2022.

Principally, a revision of the survey data is not planned for this survey year and the following years. 

Unplanned revisions: In justified cases however, a data revision will be considered. 

There are no revisions to report.

Not relevant

The source of the raw data is described with more details in the paragraphs below.

Name and short description of the sampling frame: 

Name of the sampling frame: German Census 2011

Name of the microcensus survey: Mikrozensus 2022 - Kernprogramm. 

Type of source: The source of the sample is the German Census 2011 with yearly update by building permits. Sampling is based on areas.

Survey vehicle:

Since 2021, the ICT survey is embedded in the German microcensus system.
The German microcensus is a random cluster sample covering 1% of the German population.

Survey participation:

The microcensus survey with its sub-samples (LFS, SILC, ICT) is partly mandatory and partly voluntary: the socio-demographic background variables are included in the mandatory core program of the German microcensus, as well as two ICT-related variables. The latter are the household-based variable IACC (household has got internet access) and the individual-based variable IU on internet use in the last 3 months (only the IU=1 sub-item of this question). For all other parts of the survey, participation is voluntary.

Shortcomings:

The microcensus sample (and so the included ICT sample as well) is updated on a yearly base to ensure a high degree of timeliness.

Sampling districts are distributed equally amongst their region and have the same probability to be selected. Therefore, there are no shortcomings in terms of timeliness, geographical coverage that Destatis is aware of.

The only subpopulation that is under-represented is the group of the homeless as they are not part of the sampling frame. Besides that, it is not always clear whether (new) buildings are already habitable.

Sampling design / probability design: The sampling design is a probability design.

Sampling stages: One stage cluster sample.

First stage stratified cluster sampling:

The method used to select the sampling units is stratified cluster sampling (German microcensus selection district "Auswahlbezirk"), where the German Census of 2011 served as the sampling frame. The addresses were stratified by a combination of technical (building size classes) and regional information (districts/counties or summaries of districts) as described below. Based on the stratification, addresses were clustered into artificially delimited areas (selection areas; in German called "Auswahlbezirke") that consist of around 9 dwellings or respectively 15 persons. Based on the specifications of the Census like clustering, the selected units were sampled and all sampling information was determined. Due to legislation only 20% of the sampling frame was stored. Based on the determined sampling variables, the sample districts were assigned to the microcensus and to the ICT-subsample. These variables are assigned as random numbers by permutation and are used to delimit the annual sample of the microcensus as well as the disjoint subsamples like ICT (3 sub-samples: LFS, SILC, ICT).

Stratification variables referring to building size classes:

The first stratum includes smaller buildings with 1 to 4 apartments. They are grouped into sample districts with a guideline value of 12 apartments, in the order of the house numbers within the street, if necessary also across streets. The second stratum includes medium-sized buildings with 5 to 10 apartments. These buildings each form their own selection districts. The third stratum includes buildings with 11 or more apartments. These are divided into sections with a guide size of 6 apartments. There are two more strata: Stratum N°4 includes the population in communal accommodation, which is divided into selected units with a reference size of 15 people. Another stratum N°5 serves for updating the basic selection. This annual update of the selection takes place via the reports on the construction activity statistics (so-called building permits). The new buildings registered there are divided into the three initial size classes mentioned.

Regional stratification variables:

There are 243 regional strata (districts or groups of districts), which as a rule should have at least 200 000 inhabitants. The technique of selection, i.e. the sorting, zone formation and selection per zone ensures a stratification-like effect for these regions.

Number of individuals interviewed in the household: 

All household members in the EU scope were asked to participate in the survey on voluntary basis. Thus, either one or some or all members of the "in scope" household members were interviewed, dependent on their willingness to participate. 

Longitudinal component:

Approximately 30% of the microcensus selection areas are replaced annually by new selection areas to be included in the selection (principle of partial rotation; rotation takes place at the level of the microcensus). This means that approximately 30% of the households surveyed from the previous year are excluded from the survey, while an equal proportion of households to be surveyed is included in the survey for the first time. The ICT subsample of the German microcensus is affected by this, as the other subsamples (LFS, SILC) also do. In contrast to "real" panel surveys (e.g. EU-SILC), the ICT survey is however not to be regarded as a panel survey, because no follow-up of the ICT sample households takes place from year to year. 

Annual

1) Methods used to gather data:

Multi-mode data collection (CAPI, CATI, CAWI, PAPI).

2) Short description of the survey method:

3) Variables completed from an external source: None.

A number of IT programs and tools are established and tested, among them are the IT tools for the various data collection modes (CAPI, CATI, CAWI, Paper-and-Pencil). The mode-related IT tools include already numerous consistency and plausibility controls as well as instructive content for the respondent. 

After data receipt from the Federal Statistical State Offices, the collected data are further examined and processed by the FSO. Response rates, consistency and plausibility of the data is checked comprehensively. Adequate imputation methods are applied, if necessary. 

The checked and cleaned data are then weighted by the FSO, and the weighted results are in-depth-analysed by the FSO.

Based on the finalized national data, the EU target variables are constructed by the FSO, according to the EU requirements (transmission format). Additional calculations are made (e.g. net equivalence income quintiles and other required data set content). The national data set is supplemented by the variables required for the EU. Consistency and coding of the constructed EU target variables is analysed by the FSO. The fulfilment of EU validation rules is checked.

Before providing the German data sets (via EDAMIS), Eurostat validation standards are applied/checked in order to guarantee that the data to be delivered fulfil the validity criterion requested by Eurostat transmission standards.

Data editing:

Data entry is done by the Federal Statistical State Offices. A very complex IT application ("Fachanwendung MIKIS") is used for that. MIKIS contains a multitude of plausibility and consistency checkings. It also comprises the whole administrative management (including management of the dunning procedure).

Imputation:

Deductive imputation was applied in some cases of item non-response, where appropriate (e.g. "Yes/No" questions with only "Yes" sub-items ticked and remaining sub-items missing).
In such cases, the missings were re-coded, assuming that those missings are actually "No" answers.

Methods used to impute item non-response:

Imputation of non-response was undertaken in some cases where it seemed to be appropriate to apply a deductive method.
See also 18.5.2.

Weighting: See 18.5.3

For the target indicator "Individuals having ordered goods or services for private use over the internet in the last 12 months" (individuals who ticked 'Within the last 3 months' or 'Between 3 months and a year ago' in question D1 of the 2022 model questionnaire):

Imputation rate (% of observations): 0.0%

Imputation rate (share of estimate): 0.0%

No additional issues on imputation methods.

Grossing up procedures have been applied to: Individuals and Households

Description of the weighting procedures:

Weighting is done separately for households and individuals using the same procedure, but it is applied to auxiliary variables of the respective unit. Design weights are calculated as the inverse of the inclusion probabilities. Every unit (household or person) receives the inclusion probability of the associated selection area (district; "Auswahlbezirk"). Inclusion probabilities are known from sample selection and differ between NUTS-2 regions. The design weights are adjusted by response probabilities.

Since ICT is integrated as a submodule into the compulsory Microcensus survey, the ICT gross sample can be estimated by the associated core sample. Response probabilities are calculated by logit regression. Input weights for calibration are calculated as the inverse of the product of inclusion and response probability. At the time of EU-ICT weighting, for some variables more information is available in the form of weighted totals from a bigger sample (Microcensus-core).

For the integrated system, the following two phase-procedure was implemented:

First phase: Adjustment of Microcensus core to updated population data (age, sex, nationality) by using the Generalized Regression (GREG) Estimator. The resulting weights are used for the calculation of the reference-data.

Second phase: Adjustment of ICT survey to the weighted core by using GREG-estimation, too. A lot of information is available on both household and population level for weighting of the ICT subsample. It is to outline that both phases are part of the weighting itself.

Additional steps for the calculation of response probabilities are done in advance as described.

Not relevant

Not relevant

Annexes of this report as a whole:

- German ICT household survey 2022; national language questionnaire (PAPI version)
- German ICT household survey 2022; english language questionnaire (PAPI version)

Annexes incorporated as chapter annexes in this report:

10.6. Documentation and methodology:
- Introduction to the microcensus survey
- Preparation of the microcensus operation
- Instructions for visiting the selected microcensus districts
- Instructions and Explanations on the questionnaire(s)

13.2.1. Sampling errors - indicators:
- German_ICT_household_survey_2022: relative standard errors

15.4.1 Questionnaire 2022 - mandatory questions:
- German_ICT_household_survey_2022: coverage of mandatory questions

15.4.2 Questionnaire 2022 - optional questions:
- German_ICT_household_survey_2022: coverage of optional questions