Croatian Bureau of Statistics 

Demographic and Social Statistics Directorate

Education, Culture and Information Society Statistics Department

Branimirova 19

10 000 Zagreb

The Adult Education Survey (AES) covers adults’ participation in education and training (formal - FED, non-formal - NFE and informal learning - INF). The 2022 AES focuses on people aged 18-69. The reference period for the participation in education and training is the twelve months prior to the interview.

Information available from the AES is grouped around the following topics:

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

- Classification of Learning Activities (CLA, 2016 edition)

- International Standard Classification of Education 2011 (ISCED 2011)

- Fields of Education and Training 2013 (ISCED-F)

- Classification of Occupations 2008 (ISCO-639) 

- Classification of economic activities Rev. 2 (NACE Rev. 2)

AES covers all economic sectors.

Definitions as well as the list of variables covered are available in the 2022 AES implementation manual (http://ec.europa.eu/eurostat/web/education-and-training/methodology).

The statistical units for the 2022 AES are the individuals and the non-formal learning activities.

The target population of the 2022 AES consists of individuals living in private households and aged 18 to 69 years.

People living in institutions, e.g. residential homes for the elderly, nursing homes, prisons, etc., are excluded from the survey.

The survey was conducted in four NUTS regions (Pannonian, Adriatic, Northern Croatia and City of Zagreb). No country parts were excluded from the survey.

2007 AES, 2016 AES, 2022 AES.

The field work for AES 2022 was carried out in the period from November 15, 2022 to February 13, 2023.

Not applicable.

Number, EUR.

The reference period for the participation in education and training is the twelve months prior to the interview.

The field work for AES 2022 was carried out in the period from November 15, 2022 to February 13, 2023. 

At European level:

Basic legal act: Regulation (EU) 2019/1700

Implementing act: Commission Implementing Regulation (EU) 2021/861

At national level:

Annual Implementation Plan of Statistical Activities of the Republic of Croatia for 2023 (NN, Nos 35/2023).

Not applicable.

According to the Official Statistics Act (NN, Nos 103/03., 75/9., 59/12., 13/13. and 25/20. - consolidated text), the confidentiality of all personal data on a household and its members given by respondents is guaranteed. Collected data are used solely for statistical purpose and are published in an aggregated form.

The dataset was checked using STRUVAL/CONVAL and does not contain individual identifiers and was transmitted via EDAMIS.

First and final release will be published on 7 May 2024.

Release calendar was published on web site of Croatian Bureau of Statistics - Croatian Bureau of Statistics.

Users will be informed of the date through a release on web site of Croatian Bureau of Statistics.

Adult Education Survey in Croatia as from 2022 is carried out every 6 years.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Part of the methodology is mentioned in national First Release. Besides that, the interviewers received the methodological instructions prior to the fieldwork.

Not applicable.

This national quality report.

Prior to AES fieldwork, translated questionnaire was implemented to interviewers' laptops. The questionnaire contained filters and controls to ensure the right questions are answered. The long-term interviewers of Croatian Bureau of Statistics have been trained especially for the purposes of AES. During the fieldwork, the data was immediately withdrawn to CBS server via CMS application. Later on, the data was coded and corrected via STRUVAL/CONVAL tool provided by EUROSTAT.

Overall assessment of data quality is good.

The expected user groups of the survey are policy makers at national and European level, social actors, media, researchers, students, enterprises and International organisations. However, the most important main user group would be researchers and students.

Until now we did not get any user satisfaction feedback, national First Release is in May 2024.

Final national dataset covers all variables as requested in the 2022 AES legislation.

The data completeness (rate) is 100%.

Accuracy is overall satisfying.

Accuracy in the statistical sense denotes the closeness of computations or estimates to the exact or true values. Statistics are not equal with the true values because of variability (the statistics change from implementation to implementation of the survey due to random effects) and bias (the average of the possible values of the statistics from implementation to implementation is not equal to the true value due to systematic effects). Several types of error occur during the survey process which comprises the error of the statistics (their bias and variability). The main indicator used is the coefficient of variation (CV).

Definition of coefficient of variation:

Coefficient of Variation= (Square root of the estimate of the sampling variance) / (Estimated value)

That is why sampling errors in terms of standard errors, coefficient of variation and confidence intervals were estimated for key AES indicators and they are reduced to minimum after calibration process.

Taylor linearization method was used for variance estimation by applying SAS procedure Surveymeans for calculating variance estimation. Sample design and weighting has been taken into account because it is required by mentioned SAS procedure.

See table 13.2.1 “Sampling errors - indicators for 2022 AES key statistics” in annex “HR - QR tables 2022 AES (excel)”.

See items 13.3.1 - 13.3.5 below.

Due to lack of national register, the sampling frame was taken from the Census of Population, Households and Dwellings 2021. The time lag was too long and some amount of respondents was unknown or moved from the residence.

See table 13.3.1.1 “Over-coverage - rate” in annex “HR - QR tables 2022 AES (excel)”.

Not applicable.

All efforts have been undertaken in order to reduce measurement error.

In order to reduce unit non-response, interviewers had to leave a notification containing their name and contact to set up a new timetable. If there was no response, interviewers had to visit respondents three times in different time periods.

See table 13.3.3.1 “Unit non-response - rate” in annex “HR - QR tables 2022 AES (excel)”.

See table 13.3.3.2 “Item non-response rate” in annex “HR - QR tables 2022 AES (excel)”.

The data editing process was executed in line with STRUVAL/CONVAL rules. Coding and microdata set were compatible to European requirements.

Not applicable.

No difference.

First and final release will be published on 7 May 2024.

See above.

The data collection for the survey took place between November 15, 2022 and February 2023. Processing and transmission of microdata to Eurostat took place within 7 months after the fieldwork.

See table 14.2 “Project phases - dates” in annex “HR - QR tables 2022 AES (excel)”.

Not applicable.

National implementation of 2022 AES fully followed concepts and definitions set in the legislation and in the implementation manual. Therefore, we did not have any deviations from European concepts.

See table 15.1 “Deviations from 2022 AES concepts and definitions” in annex “HR - QR tables 2022 AES (excel)”.

No additional variables related to COVID-19 were collected.

Not applicable.

Croatia did not participate in the adult education survey in 2011 but participated in the adult education survey in 2016. Difference between last two waves was at data collection mode. In AES 2016 we used only CAPI and in AES 2022 CAPI and CATI. 

Also (in line with the legislation), we have changes in target population: in 2022 AES aged 18 to 69 years and in 2016 aged 25 to 64.

See also table 15.2 “Comparability - over time” in annex “HR - QR tables 2022 AES (excel)”.

Not applicable.

Due to different reference period, coverage of learning activities and survey methodology, AES and LFS data slightly differ. In LFS the focus is on labour information for weeks prior to the interview which makes information on education of secondary importance. On the other hand, AES is exclusively dedicated to education and training issues with reference period of 12 months.

See table 15.3 “Coherence - cross-domain” in annex “HR - QR tables 2022 AES (excel)”.

Not applicable.

Not applicable.

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

Total expenditures for AES 2022 were 54,533.48 Euro. That included compensation for outsourced interviewers and excludes cost of interviewers employed in the Croatian Bureau of Statistics. Cost of preparation and processing of date are excluded as well.

Not applicable.

Not applicable.

Not applicable.

The sample for the statistical survey on adult education is based on data from the Census of Population, Households and Dwellings 2021.

The random sampling method selected a sample of 8,000 persons, whereby the implicit stratification was made, i.e. the persons within each stratum were sorted by household and country marker, with the aim of selecting the sample from each of the 21 counties, so that the final sample was representative for the whole of Croatia. Due to the strict stratification criteria, it was not possible to satisfy the requirement that only one person is selected from each household.

The sampling frame for the statistical surveys on adult education is based on data from the Census of Population, Households and Dwellings 2021. Persons who deceased in the period from 2021 to 2022, persons under the age of 18 and above the age of 69, and persons who did not have information on the highest completed school were excluded from the basic set. The set contained 2,550,092 persons from whom the sample was selected, stratified by four statistical NUTS2 regions (Adriatic, Pannonian and Northern Croatia and City of Zagreb), according to gender, age and level of education. The target population was divided into four age groups (18-24, 25-34, 35-54, 55-69) and three groups by degree of education (ISCED 1-2, ISCED 3-4, ISCED 5-8).

See also table 18.1 “Source data” in annex “HR - QR tables 2022 AES (excel)”.

Every 6 years.

The AES 2022 was a stand-alone survey for which the sample was based on data from the Census of Population, Households and Dwellings 2021. 

The survey data was collected by CAPI (computer assisted personal interview) and CATI (computer assisted telephone interview) methods interviewing individuals selected in the sample. 

See also table 18.1 “Source data” in annex “HR - QR tables 2022 AES (excel)”.

The survey data was collected by CAPI and CATI method interviewing individuals selected in the sample. The field work was carried out by trained interviewers who were entering respondent's answers which then were automatically retrieved in the Central Bureau of Statistics data base.

The questionnaire for the implementation of the research was tested via a grant.

The questionnaire (CAPI/CATI) contained logical controls and filters to ease and ensure data quality. 

Variable HHINCOME was calculated according to the set of rules from the 2022 AES Manual. Household total net monthly income was divided by equivalent household size according to OECD modified scale. HHINCOME missing values were imputed by IVEware software.

In the 2nd stage of the weighting scheme after calculation of design weights we have defined weighting groups (strata used for stratification) regarding non-response, and response rates are calculated for each of these groups. These homogenous groups are defined with 4 age groups, 3 ISCED education level groups, 2 gender and 4 NUTS2 regions. As there were 24 groups defined, multiplying 4 regions gives 96 groups within Croatia for defining non-response rates (and calculating response rates in each one). Non-response weights are inverse value of these response rates. Non-response adjustment weights were then multiplied with the design weight to compensate non-response and reduce non-response bias.

At the end, calibration was used for adjustment weighting. It has more possibilities in comparison to other techniques (e.g. linear weighting and post-stratification), such as imposing constraints on the values of the weights. So, distributions of the nine 5-year age groups, one 6-year age group, 2 gender and 4 NUTS2 regions for the subsample of respondents was calibrated to known auxiliary information from Census 2021. It reduced sampling error, effecting the ‘old’ weights, but in limits of previously defined constraints in CALMAR software. Bounded linear method was used for calibration, and lower limit was set on 0.2, and upper on 1.8.

See table 18.5.1 “Imputation - rate” in annex “HR - QR tables 2022 AES (excel)”.

Not applicable.

Not applicable.

None.