Eurostat, the statistical office of the European Union

F3: Labour market and lifelong learning

2920 Luxembourg LUXEMBOURG

The adult education survey (AES) covers adults’ participation in education and training (formal, non-formal and informal learning) and is one of the main data sources for EU lifelong learning statistics. The AES covers the resident population aged 25-64 until 2016 and 18-69 from 2022 onwards. The reference period for the participation in education and training is the twelve months prior to the interview.

The following information is available from the AES:

• Participation in formal education, non-formal education and training and informal learning (respectively labelled FED, NFE and INF)
• Volume of instruction hours
• Characteristics of the learning activities
• Reasons for participating
• Obstacles to participation
• Access to information on learning possibilities
• Employer financing and costs of learning
• Self-reported language skills

Four waves of the survey have been implemented so far (2007 AES, 2011 AES, 2016 AES and 2022 AES). The first AES – referred to as 2007 AES – was a pilot exercise and carried out on a voluntary basis in 29 countries in the EU, EFTA (European Free Trade Association) and candidate countries between 2005 and 2008. As from 2011, the AES is underpinned by a European legal act and thus carried out in all Member States on a mandatory basis. The 2011 AES was carried out in 2011/2012, the 2016 AES in 2016/2017 and the 2022 AES in 2022/2023.

Comparable data for the four waves can be found in the following folders:

• Participation in education and training (last 12 months) (trng_aes_12m0)
• Participation in informal learning (last 12 months) (trng_aes_12m4)
• Access to information on education and training (last 12 months) (trng_aes_12m1)
• Time spent on education and training (last 12 months) (trng_aes_12m2)
• Obstacles to participation in education and training (last 12 months) (trng_aes_12m3)
• Self-reported language skills (educ_lang_00)

- Classification of Learning Activities (CLA): 2006 edition for 2007 and 2011 AES and 2016 edition for 2016 and 2022 AES;

- International Standard Classification of Education (ISCED): ISCED 1997 for 2007 and 2011 AES and ISCED 2011 for 2016 and 2022 AES;

- International Standard Classification of Education - Fields of Education and Training (ISCED-F): ISCED-F 1999 for 2007 and 2011 AES and ISCED-F 2013 for 2016 and 2022 AES;

- Classification of Occupations (ISCO): ISCO-COM 88 for 2007 AES and ISCO 08 for 2011, 2016 and 2022 AES;

- Classification of economic activities (NACE): NACE Rev. 1.1 for 2007 AES and NACE Rev. 2 for 2011, 2016 and 2022 AES.

AES covers all economic sectors.

Lifelong learning encompasses all learning activities undertaken throughout life with the aim of improving knowledge, skills and competences, within personal, civic, social or employment-related perspectives. The intention or aim to learn is the critical point that distinguishes these activities from non-learning activities, such as cultural or sporting activities.

Adult learning refers to the participation of adults in lifelong learning. Adult learning usually refers to learning activities after the end of initial education.

Learning activities are any activities of an individual organised with the intention to improve his/her knowledge, skills, and competences. They are a deliberate search for knowledge, skills or competences and they are organised in some way, i.e. the learning is planned in a pattern or sequence with explicit or implicit aims. Random learning is excluded.

Three broad categories of learning activities – formal, non-formal and informal – are defined in the International Standard Classification of Education 2011 (ISCED 2011). The Classification of learning activities (CLA) provides further details, in particular for non-formal and informal learning activities.

• Formal education and training is defined as “education that is institutionalised, intentional and planned through public organisations and recognised private bodies and – in their totality – constitute the formal education system of a country. Formal education programmes are thus recognised as such by the relevant national education authorities or equivalent authorities, e.g. any other institution in cooperation with the national or sub-national education authorities. Formal education consists mostly of initial education. Vocational education, special needs education and some parts of adult education are often recognised as being part of the formal education system.” (ISCED 2011)

• Non-formal education and training is defined as “education that is institutionalised, intentional and planned by an education provider. The defining characteristic of non-formal education is that it is an addition, alternative and/or complement to formal education within the process of the lifelong learning of individuals. It is often provided to guarantee the right of access to education for all. It caters to people of all ages but does not necessarily apply a continuous pathway-structure; it may be short in duration and/or low-intensity, and it is typically provided in the form of short courses, workshops or seminars. Non-formal education mostly leads to qualifications that are not recognised as formal or equivalent to formal qualifications by the relevant national or sub-national education authorities or to no qualifications at all. Non-formal education can cover programmes contributing to adult and youth literacy and education for out-of-school children, as well as programmes on life skills, work skills, and social or cultural development.” (ISCED 2011) In short, non-formal education and training covers institutionalised taught learning activities outside the formal education system. For the AES, four types of typical non-formal learning activities are singled out (these categories are not detailed in the online tables):
  
  • Courses
  • Workshops or seminars
  • Guided on-the-job training (planned periods of education, instruction or training directly at the workplace, organised by the employer with the aid of an instructor)
  • Private lessons

• Informal learning is defined as “forms of learning that are intentional or deliberate, but are not institutionalised. It is consequently less organized and less structured than either formal or non-formal education. Informal learning may include learning activities that occur in the family, workplace, local community and daily life, on a self-directed, family-directed or socially-directed basis”. (ISCED 2011)
  

One fundamental criterion to distinguish formal and non-formal education and training from informal learning is that for the latter, the learning activities are not institutionalised. Institutionalised learning activities occur where there is “an organisation [which] provides structured educational arrangements, such as student-teacher relationships and/or interactions, that are specially designed for education and training”. (ISCED 2011).

The participation rate in education and training covers participation in formal and non-formal education and training. Participation of adults in education and training is a measure of lifelong learning. The main indicator for participation in education and training refers to the age group 25-64.

Job-related non-formal education and training: the respondent takes part in the non-formal education and training activity in order to obtain knowledge and/or learn new skills needed for a current or future job, to increase earnings, to improve job and/or career opportunities in a current or another field and generally to improve his/her opportunities for advancement and promotion.

Employer-sponsored job-related non-formal education and training: all job-related non-formal education and training activities paid at least partially by the employer and/or done during paid working hours.

The educational attainment level of an individual is the highest ISCED (International Standard Classification of Education) level successfully completed, the successful completion of an education programme being validated by a recognised qualification, i.e. a qualification officially recognised by the relevant national education authorities or recognised as equivalent to another qualification of formal education. In countries where education programmes, in particular those belonging to ISCED levels 1 and 2, do not lead to a qualification the criterion of full attendance of the programme and normally gaining access to a higher level of education may have to be used instead. When determining the highest level, both general and vocational education should be taken into consideration. The ISCED definition of education includes training.

Three levels of education are distinguished in the tables:

• Less than primary, primary and lower secondary education: this aggregate refers to levels 0, 1 and 2 of the ISCED 2011 (online code ED0-2). Data up to 2011 refer to ISCED 1997 levels 0, 1 and 2 but also include level 3C short (educational attainment from ISCED level 3 programmes of less than two years).
• Upper secondary and post-secondary non-tertiary education: this aggregate corresponds to ISCED 2011 levels 3 and 4 (online code ED3_4). ISCED 2011 level 3 programmes of partial level completion are considered within ISCED level 3. Data up to 2011 refer to ISCED 1997 levels 3C long, 3A, 3B and 4.
• Upper secondary and post-secondary non-tertiary education - general: this aggregate corresponds to ISCED 2011 levels 34 and 44 (online code ED3_4GEN). Data are available from 2022 onwards.
• Upper secondary and post-secondary non-tertiary education - vocational: this aggregate corresponds to ISCED 2011 levels 35 and 45 (online code ED3_4VOC). Data are available from 2022 onwards.
• Tertiary education: this aggregate covers ISCED 2011 levels 5, 6, 7 and 8 (short-cycle tertiary education, bachelor's or equivalent level, master's or equivalent level, doctoral or equivalent level, online code ED5-8 ‘tertiary education’). Data up to 2011 refer to ISCED 1997 levels 5 and 6.

Degree of urbanisation:

Three types of areas are defined based on population density:

• cities: at least 50% of the population lives in urban centres;
• towns and suburbs: at least 50% of the population lives in urban clusters and less than 50% of the population lives in urban centres;
• rural areas: at least 50% of the population lives in rural grid cells.
  

Occupation:

Type of occupation defined according to the ISCO. Four categories of persons employed are distinguished in the tables:

• Managers, professionals, technicians and associate professionals;
• Clerical support workers, service and sales workers;
• Skilled manual workers;
• Elementary occupations.

Level of self-reported knowledge of a foreign language:

The knowledge of foreign languages is self-reported by the respondents and is categorised in three levels. The labels displayed in the tables of the domain educ_lang_00 correspond to the following self-reported knowledge:

• Basic: I can understand and use the most common everyday expressions. I use the language in relation to familiar things and situations;
• Good: I can understand the essential of clear language and produce simple text. I can describe experiences and events and communicate fairly fluently;
• Proficient: I can understand a wide range of demanding texts and use the language flexibly. I master the language almost completely.

In the 2007, 2016 and 2022 AES, a fourth 'very basic' level was included in the questionnaire: 'I only understand and can use a few words'. In the online tables that fourth category is included in the category 'basic'.

Individuals, non-formal learning activities

Up to 2016, the AES covers the resident population aged 25 to 64. Data for several countries refer to those living in private households only. The 2022 AES covers the resident population aged 18-69 living in private households.

2016 data for IE exclude persons in continuous fulltime education.

2007 AES: Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Italy, Latvia, Lithuania, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Norway, Switzerland, the United Kingdom and Türkiye.

2011 AES: Austria, Belgium, Bulgaria, Cyprus, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Norway, Switzerland, the United Kingdom, Serbia and Türkiye.

2016 AES: Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Norway, Switzerland, the United Kingdom, Serbia and Türkiye. Albania, Bosnia and Herzegovina and North Macedonia also participated in the 2016 round.

2022 AES: Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Iceland, Norway, Switzerland, Serbia and Türkiye. Albania and North Macedonia also participated in the 2022 round.

2007, 2011 and 2016 data for FR refer to Metropolitan France only.

2011 and 2016 data for UK refer to England only.

Depending on the country, the 2007 AES was carried out between 2005 and 2008.

The 2011 AES was carried out between July 2011 and June 2012 (in Finland until December 2012).

The 2016 AES was carried out between July 2016 and March 2017 (in Finland until June 2017 and in Ireland until January 2018).

The 2022 AES was carried out between July 2022 and March 2023 (in Malta until mid-April 2022, in Spain from April to July 2023, and in Switzerland already from April to June 2021).

Details of the national data collection periods are available in the annexes.

Not applicable.

2007 AES: gentlemen’s agreement.

2011 and 2016 AES:

Basic legal act

Regulation (EC) No 452/2008 concerning the production and development of statistics on education and lifelong learning (http://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1416236812785&uri=CELEX:32008R0452)

Implementing regulations

2011 AES

Commission Regulation (EU) No 823/2010 as regards statistics on the participation of adults in lifelong learning (http://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1415896180907&uri=CELEX:32010R0823)

2016 AES

Commission Regulation (EU) No 1175/2014 as regards statistics on the participation of adults in lifelong learning (http://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1415896180907&uri=CELEX:32014R1175) 

2022 AES:

Basic legal act

Regulation (EU) 2019/1700 on a common framework for European statistics relating to persons and households, based on data at individual level collected from samples (https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=uriserv:OJ.LI.2019.261.01.0001.01.ENG)

Implementing regulation

Commission Implementing Regulation (EU) 2021/861 on the technical items of the data set and establishing the technical formats for transmission of information on the organisation of a sample survey in the education and training domain (https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32021R0861&qid=1622465169193)

For further information see Eurostat's website page on education and training statistics legislation.

Not applicable.

Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

AES microdata as received by Eurostat from the national statistical institutes do not contain any administrative information such as names or addresses that would allow direct identification. Access to this microdata is nevertheless strictly controlled and limited to specified Eurostat staff.

For data published in the online database, confidentiality/reliability thresholds are applied. These thresholds determine the size of the sample used for computing results below which data are either not published or published with a flag. Data are either blanked or flagged if they are below the limits.

The rules are as follows:

• An estimate should not be published if it is based on less than 20 sample observations or if the non-response for the item concerned exceeds 50%.
• An estimate should be published with a flag (‘u’ for low reliability) if it is based on 20 to 49 sample observations or if non-response for the item concerned exceeds 20% and is lower or equal to 50%.
• An estimate shall be published normally when based on 50 or more sample observations and the item's non-response does not exceed 20%.

Under specific conditions, researchers may have access to anonymised microdata. For further information see access to microdata.

There is no specific release calendar. Results are published approximately 9-12 months after the end of the data collection period.

Not applicable.

In line with the Community legal framework and the European Statistics Code of Practice Eurostat disseminates European statistics on Eurostat's website (see item 10 - Accessibility and clarity) respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably. The detailed arrangements are governed by the Eurostat protocol on impartial access to Eurostat data for users.

Ad-hoc news releases when required.

See the list of publications related to AES on CIRCABC.

Statistics based on AES can be found in the following folders of the online database (see also 3.1 above):

• Participation in education and training (last 12 months) (trng_aes_12m0)
• Participation in informal learning (last 12 months) (trng_aes_12m4)
• Access to information on education and training (last 12 months) (trng_aes_12m1)
• Time spent on education and training (last 12 months) (trng_aes_12m2)
• Obstacles to participation in education and training (last 12 months) (trng_aes_12m3)
• Self-reported language skills (educ_lang_00)

Please consult Eurostat database.

2007, 2011 and 2016 AES anonymised microdata are accessible for researchers; see access to microdata for further details. 2022 AES anonymised microdata are planned to become available in Q2/2024.

See: http://ec.europa.eu/eurostat

For a detailed description of methods and concepts used, as well as for other documents related to the AES, please consult the reference documents related to 2007, 2011, 2016 and 2022 AES on CIRCABC.

See item 10.6. above.

The quality of the AES is ensured through specific requirements set in the regulations for the 2011, 2016 and 2022 AES and is also reflected through the use of harmonised definitions and concepts. Specific recommendations to help countries properly collecting the expected data are also available through a set of methodological documents and guidelines (see item 10.6. Documentation on methodology). The quality is discussed in working groups (such as the Education and Training Statistics working group), workshops and seminars within the European Statistical System (ESS).

2007 AES: while based on a gentlemen’s agreement a common EU framework including a standard questionnaire, tools and quality reporting was agreed for 2007 AES and was largely followed by the participating countries.

2011 AES: the content of the survey, i.e. the variables to be delivered to Eurostat, is precisely determined in Commission Regulation (EU) No 823/2010. Sampling and precision requirements have been set up within this regulation. Detailed guidelines were agreed in the 2011 AES manual.

2016 AES: the content of the survey, i.e. the variables to be delivered to Eurostat, is precisely determined in Commission Regulation (EU) No 1175/2014. Sampling and precision requirements have been set up within this regulation. Detailed guidelines were agreed in the 2016 AES manual.

2022 AES: the content of the survey, i.e. the variables to be delivered to Eurostat, is precisely determined in Commission Implementing Regulation (EU) No 2021/861. Sampling requirements have been set up within this regulation, while precision requirements are set up within Regulation (EU) 2019/1700. Detailed guidelines were agreed in the 2022 AES manual.

AES statistics are considered to be of good quality thanks to a harmonised production process (as described in item 11.1. above). However, like any other survey, it is based on a sample of the population meaning that results are subject to the usual statistical errors of measurement.

National quality reports provide users with basic information on quality at national level and give further explanations about the possible weaknesses of the sampling methods used at national level and of the final national effective sample of the survey.

National quality reports for the 2022 AES are expected to become available in Q1/Q2 2024.

For national quality reports for the 2016 AES see 'national metadata' at the top of this document.

2007 and 2011 AES quality reports are available on CIRCABC.

For 2016, summary tables based on the national quality reports are available on CIRCABC.

See also the reference documents related to 2007, 2011 and 2016 AES on CIRCABC.

Statistics based on AES support monitoring adult participation in lifelong learning at EU level by providing detailed results on the participation (participation rates, reasons for participating, characteristics of the learning activities, outcomes, etc.) and the non-participation (obstacles to participation).

The participation of adults in lifelong learning holds a high profile both on the European policy agenda – particularly with a view to the "European pillar of social rights", the "European skills agenda" as well as the "Strategic framework for European cooperation in education and training towards the European Education Area and beyond" – and at national level where many countries put in place specific frameworks for enhancing the skills of the adult population in order to increase their employability.

Therefore, high quality data on participation in formal and non-formal education and training are crucial to better underpin the actions taken under the various initiatives targeting lifelong learning.

There is no satisfaction survey targeted at adult learning statistics users. Whenever it comes to defining the content of the adult education survey and the type of results that are disseminated by Eurostat on its website, Eurostat consults stakeholders as much as possible to get their opinion and to satisfy their needs in terms of data availability.

The data sent by participating countries to Eurostat are overall complete and match the requirements set out in the legislation (2011, 2016 and 2022 AES) or gentlemen’s agreement (2007 AES) respectively.

Nevertheless, some national datasets are not always fully matching the expected format because some content is missing. In those cases the data disseminated are displayed as ‘not available’ (‘:’). This can be explained either because the country could not implement the variable for some reason or because the variable once collected was of very poor quality due to various factors (too high non-response rate or errors in the production process for instance).

The overall accuracy of the AES is considered as high. The sampling designs are chosen by countries according to EU recommendations. While designing their sample, countries must make sure that the provision of data will comply with the precision requirements set out in the regulations (2011, 2016 and 2022 AES).

Most of the national statistical authorities use multi-staged stratified random sample design, especially those that do not have central population registers available.

Regardless of the sampling method or which age groups are interviewed, the datasets sent to Eurostat by countries are representative for the population aged 25-64 up to 2016 and 18-69 from 2022 onwards.

As the results are based on a sample of population they are subject to the usual types of errors associated with sampling techniques and interviews. Sampling errors, non-sampling errors, measurement errors, processing errors and non-response are calculated for each country and documented in the quality reports (see item 11.2).

The participating countries provide Eurostat with an estimate of the relative standard error of the indicator on participation of adults in non-formal education and training (up to 2016 for the age group 25-64, as from 2022 for the age group 25-69). The relative standard error can also be expressed as a confidence interval, i.e. the range of values that in 95% of the cases would capture the true value in the population.

Since 2022, with the extension of the age coverage to 18-69, countries also provide an estimate of the relative standard error for the participation in formal education and training of those aged 18-24.

The estimates and confidence limits are calculated by each country and documented in the quality reports (see item 11.2).

There are four types of commonly reported non-sampling errors. The four types described below are mentioned in the quality reports (see item 11.2).

a) Coverage errors:

Non-existent and inhabited dwellings or population no longer living in the country are the main causes of over-coverage, especially for countries using previous censuses as sampling frames.

Under-coverage problems are caused by the time-lag in registering new residents or newly constructed dwellings.

b) Measurement errors:

Measurement errors cannot be estimated as such as they rather concern a potential bias due to the subjective approach of questions and/or the use of proxies to answer the questionnaire.

Number of proxy interviews and existence of pilot testing of the questionnaire (which enables to avoid a too large subjective interpretation of the questions) are given in the quality reports.

c) Processing errors:

Between the data collection and the beginning of statistical analysis for the production of statistics, data must undergo a certain processing: coding, data entry, data editing, imputation, etc.

The corresponding techniques used at national level are mentioned in the quality reports.

d) Non-response errors:

Non-response can be due to a failure in contacting the individual, a refusal or another reason (rejected interviews, inability to respond, etc.).

The detailed non-response rates, broken-down by type of non-response, are given in the quality reports.

AES data are released approximately within 9-12 months following the end of the data collection.

The regulations require that the data are sent to Eurostat within 6 months after the end of the national data collection period. For the 2011, 2016 and 2022 AES this deadline was met by the majority of countries.

Comparability across countries is considered as high. Comparability across countries is achieved through the regulations (2011, 2016 and 2022 AES) and the AES manuals ensuring harmonisation of methodology, concepts and definitions for all EU Member States, EFTA, candidate and potential candidate countries participating in the survey. However, perfect comparability of statistical data across countries is difficult to achieve even for a survey carried out at the same time in all concerned countries and using the same questionnaire and a single method of recording.

Comparability for the statistics on participation in education and training is ensured by:

(a) the recording of the same set of characteristics of learning activities in each country;

(b) a close correspondence between the EU list of questions and the national questionnaires;

(c) the use of the same definitions for all countries;

(d) the use of common classifications (e.g. CLA for the type of learning activities, ISCED for the level of education);

(e) the data being centrally processed by Eurostat.

Each country has the responsibility to ensure that the national survey provides data that are compatible with the EU definitions and of the same quality. However, in spite of the close coordination between the national statistical authorities and Eurostat, there inevitably remain some differences in the survey from country to country.

Not all 2007, 2011, 2016 and 2022 AES results are directly comparable and therefore cannot be used to comment on the evolution of adult participation in lifelong learning between 2007 and 2022 due to methodological changes.

For instance the definitions of some variables have changed or, more frequently, the answer categories proposed to the individuals when they are asked a question have been rephrased or improved. On the other hand, new variables have been added from one wave to another, while some others have been dropped.

Reasons for changes were to improve the survey in view of experiences of the previous wave or to meet new user or political requirements.

For further details on breaks in series see the quality reports related to 2007, 2011 and 2016 AES (see item 11.2) as well as the detailed information on the calculation of indicators and information on flags available in the annexes.

2022 AES: CH participates in AES on a voluntary basis. The fieldwork took place in 2021 and some variables were omitted. Only one random activity was selected for the details on non-formal education activities.

Results from the adult education survey are not directly comparable with educational variables coming from the EU labour force survey (LFS) which collects quarterly data on participation in formal and non-formal education and training with a reference period of 4 weeks. Main differences are the reference period (12 months in AES, 4 weeks in LFS), the coverage of non-formal education (LFS does not cover guided on-the-job training) and the overall purpose and design of the surveys (especially the fact that proxies are possible and common when collecting the LFS data while they are not recommended and almost never used in most countries when collecting the AES data), and more generally that the LFS is a survey on labour market participation while the AES is a survey on participation in learning activities).

The EU-LFS ad-hoc module on 'lifelong learning' as carried out in 2003 (trng_aes_005h) is not directly comparable with the AES information because of the survey methodology.

More information is available on CIRCABC.

Results based on AES for a given year are based on the same microdata and results are calculated using the same estimation methods, therefore the data are internally coherent.

AES data are revised when major errors are identified in the data delivered or in their processing, but there is no specific revision policy (no revision planned ahead).

Eurostat publishes the most up-to-date information available at EU-level based on the most up-to-date national microdata available and recalculates the EU aggregates whenever a national dataset has been changed.

Not available.

2007, 2011, 2016 and 2022 AES (see item 3.1).

The AES is a random sample survey of the resident population aged 25-64 up to 2016 and 18-69 from 2022 onwards.

From the 2011 AES participation in the survey is compulsory for EU Member States (see item 6.1).

2011 AES: the effective sample sizes ranged between 2 404 for the smallest sample to 22 522 for the biggest.

2016 AES: the effective sample sizes ranged between 1 963 for the smallest sample to 20 577 for the biggest.

2022 AES: the effective sample sizes ranged between 2 448 for the smallest sample to 33 790 for the biggest.

This is due to both national requirements and a big variance in the number of inhabitants per country.

Up to 2016: every five years. From 2016: every six years.

Data collection methods were designed in collaboration with the participating countries and Eurostat through the work of AES task forces and other working groups.

Data validation is done by national statistical institutes or other statistical authorities that are responsible for the survey.

Eurostat carries out quality checks, mainly on the coherence of the information provided.

The checks have two dimensions:

- a first check is made on the microdata received using a software testing the integrity of the data received, where both the structure and the logic of the data are verified. In case of doubts, a list of possible problems is sent back to countries for a data revision;

- a second check is then made on aggregated data to verify the results of key indicators in order to get feedback from countries to know whether the results computed from the microdata are correct.

Estimates for EU and euro area aggregates are calculated as the population-weighted arithmetic average of national data.

The totals are compiled based on the available countries

• 2007 without Ireland and Luxembourg
• 2011 without Croatia

For some tables, data for some countries are not complete (see item 12.3). In such cases, EU and euro area aggregates are shown based on available countries as long as 80% of the EU population is covered.

No adjustments are made apart from those possibly made by countries prior to sending their data.