Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.
National Statistics Office (NSO), Xatt Lascaris, Lascaris, Valletta, VLT 2000, Malta
1.6. Contact email address
Confidential because of GDPR
1.7. Contact phone number
Confidential because of GDPR
1.8. Contact fax number
Confidential because of GDPR
2.1. Metadata last certified
3 April 2023
2.2. Metadata last posted
3 April 2023
2.3. Metadata last update
3 April 2023
3.1. Data description
The Continuing Vocational Training Survey (CVTS) collects information on enterprises’ investment in the continuing vocational training of their staff. Continuing vocational training (CVT) refers to education or training measures or activities which are financed in total or at least partly by the enterprise (directly or indirectly). Part financing could include the use of work-time for the training activity as well as financing of training equipment.
Information available from the CVTS is grouped around the following topics:
- Provision of CVT courses and other forms of CVT (training/non-training enterprises)
- CVT strategies
- Participants in CVT courses
- Costs of CVT courses
- Time spent in CVT courses
- Characteristics of CVT courses
- Assessment of CVT activities
The CVTS also collects some information on initial vocational training (IVT).
The main groupings for enterprises are by economic activity (NACE), size group and training/non-training enterprises.
3.3. Coverage - sector
No deviations from the coverage listed in the manual. All economic activities listed as per guidelines were followed (i.e. sections B to N and R to S of NACE Rev. 2).
There were no deviations from the definition of statistical unit as per manual. The enterprise definition is compliant with Council Regulation (EEC) No 696/93. More complex enterprise will be available from next year.
3.6. Statistical population
The enterprises covered were with 10 or more persons employed belonging to the certain NACE categories as per manual (see also 3.3).
The target population applied followed guidelines, with no deviations. It contains 2952 enterprises.
Variable A2tot (persons employed) refers to 31 December 2020.
3.7. Reference area
No geographical deviations from Eurostat guidelines.
3.8. Coverage - Time
Data is available for all CVTS cycles since its first implementation in 1999.
3.9. Base period
Not applicable.
Number, EUR.
The reference year for CVTS 6 is the calendar year 2020.
6.1. Institutional Mandate - legal acts and other agreements
At European level:
Basic legal act: Regulation (EC) No 1552/2005 of the European Parliament and the Council
Implementing act: Commission Regulation (EU) No 1153/2014, amending Commission Regulation (EC) No 198/2006
At national level:
The Malta Statistics Authority Act, 2000. The Act empowers the NSO to collect, compile, extract and release official statistics related to demographic, social, environment, economic and general activities and conditions of Malta.
As a member of the European Union (EU), Malta has to observe the Council Regulation (EC) No. 223/2009 dealing with the collection and dissemination of official statistics. Moreover, all council regulations dealing with the collection of official statistics need to be observed by the NSO.
6.2. Institutional Mandate - data sharing
Not applicable.
7.1. Confidentiality - policy
At National level:
The NSO requests information for the compilation of official statistics according to the articles of the MSA Act – Cap. 422 and the Data Protection Act – Cap. 586 of the Laws of Malta implementing the General Data Protection Regulations (GDPR).
Article 40 of the MSA Act stipulates the restrictions on the use of information while Article 41 stipulates the prohibition of disclosure of information. Furthermore, Section IX of the Act (Offences and Penalties) lays down the measures to be taken in case of unlawful exercise of any officer of statistics regarding confidentiality of data.
Since its inception, the NSO has always assured that all data collected remains confidential and that it is used for statistical purposes only according to the articles and derogations stipulated in the laws quoted above. The Office is obliged to protect the identity of data providers and refrain from divulging any data to third parties that might lead to the identification of persons or entities.
During 2009, the NSO has set up a Statistical Disclosure Committee to ensure that statistical confidentiality is observed, especially when requests for microdata are received.
Upon employment, all NSO employees are informed of the rules and duties pertaining to confidential information and its treatment. In line with stipulations of the MSA Act, before commencing work, every employee is required to take an oath of secrecy whose text is included in the same Act.
An internal policy on anonymisation and pseudo-anonymisation is in place to ascertain that adequate methods are used for the protection of data which the office collects and shares with the public in its capacity as the National Statistics Office. The policy is meant to safeguard confidentiality of both personal and business data entrusted to the NSO. The document provides guidance for all NSO employees who process data on a daily basis as to how anonymisation and pseudo-anonymisation methods should be applied. The policy applies to all confidential, restricted and internal information, regardless of form (paper or electronic documents, applications and databases) that is received, processed, stored and disseminated by the NSO.
At European level:
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.
7.2. Confidentiality - data treatment
Data on provision of CVT courses is only published at the discretion of the Office. If disseminated, the NSO guarantees confidentiality of data by publishing only aggregated and un-identifiable data. Furthermore, in order to avoid publication of figures which are statistically unreliable, NSO applies the (:) flag for estimates with less than 20 un-weighted counts and (u) for estimates having between 20-50 responding counts.
8.1. Release calendar
No publication yet, planned in 2023.
8.2. Release calendar access
No publication yet.
8.3. Release policy - user access
An internal policy on dissemination is in place to govern the dissemination of official statistics in an impartial, independent and timely manner, making them available simultaneously to all users.
The NSO’s primary channel for the dissemination of official statistics is the NSO website. Tailored requests for statistical information may also be submitted through the NSO website.
Every 5 years.
10.1. Dissemination format - News release
Not applicable.
10.2. Dissemination format - Publications
One publication per data collection.
10.3. Dissemination format - online database
No online database currently available for publication.
10.3.1. Data tables - consultations
Not applicable.
10.4. Dissemination format - microdata access
Anonymised micro-data is transmitted to Eurostat in order to enable the compilation of customised reports by Eurostat for international users.
10.5. Dissemination format - other
Not available.
10.5.1. Metadata - consultations
Not applicable.
10.6. Documentation on methodology
This quality report and the publication planned in 2023.
10.6.1. Metadata completeness - rate
Not applicable.
10.7. Quality management - documentation
All methods used for data analysis were as recommended by Eurostat manuals.
11.1. Quality assurance
Audits and data validation by the team were made directly with the enterprise either by e-mail or by phone. All questionnaires were checked by the team to ensure that minimum questions are skipped by enterprises and go-tos are followed correctly.
11.2. Quality management - assessment
Overall, the survey output was of good quality. Whenever some variables had issues of under reporting, this was treated through imputation methods which were checked not to cause any biases and not deviate from the provided figures by the enterprises. Clarifications were done with the enterprise if some figures were found to be not plausible.
12.1. Relevance - User Needs
Policy makers at European level (e.g. European Commission, European Parliament, other European agencies): European Commission: CVT strategies (objectives, decision process), CVT characteristics (internal/external courses, other forms, subjects, providers), CVT volume (participants, hours), CVT costs, CVT quality, outcomes and difficulties and IVT.
Policy makers at national level (e.g. ministries): Ministry for Business: CVT strategies (objectives, decision process), CVT characteristics (internal/external courses, other forms, subjects, providers), CVT volume (participants, hours), CVT costs, CVT quality, outcomes and difficulties and IVT.
Social actors (e.g. employers' associations, trade unions): Trade Unions: CVT characteristics (internal/external courses, other forms, subjects, providers) and CVT costs.
Enterprises: for own market research activities or for consultancy services in the information sector: CVT costs, CVT quality, outcomes and difficulties and IVT.
Researchers and students: for research.
12.2. Relevance - User Satisfaction
No measure is available for the CVTS. The most recent general NSO user satisfaction survey is dated 7th July 2022.
12.3. Completeness
The final dataset covers all NACE sectors, enterprise size groups and variables as requested in the CVTS 6 legislation.
12.3.1. Data completeness - rate
Not applicable.
13.1. Accuracy - overall
A sample coverage of all local enterprises in operation within the employment and NACE strata defined in point 3.3 are provided by Business Register Unit within NSO. Any enterprises which are not operating or have less than 10 persons employed are excluded from the target population. Consistency checks are carried out with data provided from previous 5 years.
For CVTS 2020, the team managed to obtain most of the answers of the key and core variables of the questionnaire. The non-response of the other items were filled in by using hot deck imputation. Overall, item non-response of each variable was very low. No variable had an item non-response greater than 10%.
Simple random sampling is purposely used to ensure that the sample is uniformly distributed among different strata and to ensure that the population is accurately represented.
13.2. Sampling error
The sampling method used was that of simple random sampling after stratifying the population. The strata considered where according to the NACE and size categories as required by the legislation. The sampling counts per stratum for this survey are based on the proportion of training enterprises in each stratum which are taken from the previous CVTS, in this case from CVTS reference year 2015.
During the collection process, each strata was checked to ensure enough coverage and eliminate under-coverage as much as possible. The team tried to collect questionnaires which were under represented in certain strata and ensured to obtain nearly all the big enterprises (for those employing 250+ persons employed).
The sample data are firstly weighted based on the NACE and size classes. Then an extensive outlier detection process was conducted using percentiles. Those data which were deemed to be identified as ‘outliers’ were given a weight of 1. The rest of the data in the NACE and size classes were arranged accordingly so that all the population is represented. Afterwards, this adjusted weight was then calibrated to the total employment of the population coming from the BR 2020.
13.2.1. Sampling error - indicators
See table 13.2.1 "Sampling errors - indicators" in annex "MT - QR tables CVTS 2020 (excel)".
13.3. Non-sampling error
See 13.3.1 - 13.3.5.
13.3.1. Coverage error
A stratified sampling mechanism was used on the Business Register for the respective survey year, where the strata comprised of the NACE and Size Categories according to the legislation.
See table 13.3.1 "Coverage error" in annex "MT - QR tables CVTS 2020 (excel)".
13.3.1.1. Over-coverage - rate
See table 13.3.1.1 "Over-coverage - rate" in annex "MT - QR tables CVTS 2020 (excel)".
13.3.1.2. Common units - proportion
Not applicable.
13.3.2. Measurement error
Measurement error testing the questionnaire before sending: The questionnaire was analysed for potential loopholes that may have existed. The team checked every go-tos and anomalies that may have been present in the previous version of the CVTS 2020 questionnaire. Interviewers were provided a briefing which explained all the concepts and questions of the questionnaire.
Measurement errors by interviewers: The possibility of not understanding or misinterpreting the respondents' answers and typing errors. Only 28% of the received questionnaires were followed-up by the interviewers, i.e. interviewers contacted the businesses through emails and phone calls to ensure that questionnaires were filled in. To reduce this error a detailed presentation and briefing was presented to the interviewers so that they are aware of the concepts of CVTS. After the work of the interviewers, the team checked again the questionnaires received for data validation and if some questions had some missing answers, these were asked again to the enterprise.
Measurement errors by respondents: This includes the possibility of not understanding the questions, not knowing the answers, not interested in the survey, not willing to provide information on certain question and finding the questionnaire too difficult to understand. The enterprises were given a contact number and the team helped the enterprises with any queries they may have had.
Efforts done to reduce measurement errors: This was achieved by analysing the CVTS 2020 collected data against ready available sources within NSO to ensure data quality and reduce errors.
13.3.3. Non response error
Companies which did not send their feedback by the deadlines which were outlined in the letter sent to them together with the survey, were sent a number of reminders in order to remind them about the pending survey. A reminder was sent every week. Non-responding companies for certain stratas and NACE groups, were re-contacted by either email or telephone. No high-level of item non-response which exceeds the 10% mark was seen, especially in core and key variables. All the questionnaires were data validated by the team and if the questionnaire had more than three missing questions, follow up by the team was done. This ensured that very little non-response will be present. The team ensured that all the key and core variables were answered. If the key or core variables were still missing, the team followed up with the company until an answer was provided. For those items (not core or key variables) which even after chasing the companies were still left unanswered, hot deck imputation was performed. Certain sensitive questions relating to expenditure and costs where re-validated in order to minimise the risk of under-reporting.
13.3.3.1. Unit non-response - rate
See table 13.3.3.1 "Unit non-response - rate" in annex "MT - QR tables CVTS 2020 (excel)".
13.3.3.2. Item non-response - rate
See table 13.3.3.2 "Item non-response - rate" in annex "MT - QR tables CVTS 2020 (excel)".
13.3.4. Processing error
The programme collected the questionnaires which had been filled-in by the businesses and received by email (CAWI). The built in validations were used to check that no sections were skipped etc. The main issues with regards to processing errors were that of auditing of data entry in order to minimise coding errors. The data entry process was conducted using a programme designed by the IT unit within the NSO. It contained a number of built-in validations in order to highlight any values that appeared unusual. For such instances, the enterprise was re-contacted for verification. While analysing the data, tools and imputation methods used were chosen such that past information (whenever available, even from other sources) was kept relevant and not create any biases. Additional coherence checks with other statistical sources were also made. Main errors detected in the post-data collection process were namely that companies did not follow the go-to as indicated in the questionnaire and consequently some variables were found to be missing. Due to this, the companies needed to be re-contacted in order to obtain the missing information.
13.3.5. Model assumption error
Not applicable.
14.1. Timeliness
Calendar year refers to 2020. CVTS asks about the enterprise’s current status in the reference year 2020 while certain particular questions may also ask for information regarding the previous year.
14.1.1. Time lag - first result
Date of publication is scheduled for Q4 2022.
14.1.2. Time lag - final result
Date of publication is scheduled for Q4 2022. Publication of results on Eurostat website was in October 2022.
14.2. Punctuality
Countries should transmit data to Eurostat no later than 18 months after the end of the reference year. Data was sent to Eurostat on 23rd June 2022.
See table 14.2 "Project phases - dates" in annex "MT - QR tables CVTS 2020 (excel)".
14.2.1. Punctuality - delivery and publication
Not applicable.
15.1. Comparability - geographical
No geographical comparability problems.
No additional variables related to COVID-19 were collected.
See table 15.1 "Comparability - geographical" in annex "MT - QR tables CVTS 2020 (excel)".
15.1.1. Asymmetry for mirror flow statistics - coefficient
Not applicable.
15.2. Comparability - over time
No additional changes were made between CVTS 6 and CVTS 5.
See table 15.2 "Comparability - over time" in annex "MT - QR tables CVTS 2020 (excel)".
15.2.1. Length of comparable time series
Not applicable.
15.3. Coherence - cross domain
See table 15.3 "Coherence - cross-domain" in annex "MT - QR tables CVTS 2020 (excel)".
The number of persons employed of structural business statistics (SBS) is the average of the year whilst the number of persons employed of CVTS 2020 is as at end of year (31st December 2020). Moreover, SBS does not cover Sections K, R and S.
15.3.1. Coherence - sub annual and annual statistics
Not applicable.
15.3.2. Coherence - National Accounts
Not applicable.
15.4. Coherence - internal
CVTS results for a given reference year are based on the same microdata and results are calculated using the same estimation methods, therefore the data are internally coherent. Enterprises' background information is compared to other internal data sources. These include Job Vacancies Statistics (JVS), Structural Business Statistics (SBS) and the Business Register (BR).
One reduced (30 hours a week) principal statistician and one part-time clerk were involved in the data collection of CVTS data. The team carried out the data validation and the analysis of results. Other major costs associated with this survey mainly included payments to interviewers (for the second round of non-response surveys). Another full-time IT officer (working on the project for approximately two weeks) was involved in producing the CVTS 2020 application programme where completed surveys were uploaded.
The bare minimum of burden on respondents was exercised in order to collect correct and useable information. Whenever possible, we sent the questionnaire by e-mail in order to reduce costs associated with paper and mail. Nevertheless, if the enterprise did not have any valid e-mail address or the email was undelivered, the questionnaire was sent by post. The majority of the enterprise answered the questionnaire by email.
17.1. Data revision - policy
Not applicable.
17.2. Data revision - practice
Not applicable.
17.2.1. Data revision - average size
Not applicable.
18.1. Source data
See table 18.1 "Source data and data collection" in annex "MT - QR tables CVTS 2020 (excel)".
18.2. Frequency of data collection
Every 5 years.
18.3. Data collection
See also table 18.1 "Source data and data collection" in annex "MT - QR tables CVTS 2020 (excel)".
18.4. Data validation
All questionnaires were checked, validated and audited by the team.
18.5. Data compilation
From the data entry programme, an extract of data was issued for basic data analysis following the rules established in the manual. Any queries regarding the received dataset was made at this stage, while additional imputations were tagged for later analysis. As already explained outlier detection was implemented and those companies which deemed to be outliers were given a weight of 1. The other weights in the same NACE and Employment size strata were adjusted accordingly to reflect the population. The steps in the weighting system were the following: first the initial weight was extracted, then another weight referred to as 'adjusted weight' was created. This 'adjusted weight' took into consideration the outlier detection procedure and each outlier was given a weight of 1 and other weights in the same strata were adjusted accordingly. The third and final weight disseminated is the calibrated weight which takes the 'adjusted weight' and calibrates it against the total employment of the population. In case of non-response of other variables (nor core or key), hot deck imputation was implemented so that no missing variables were transmitted.
18.5.1. Imputation - rate
See table 18.5.1 "Imputation - rate" in annex "MT - QR tables CVTS 2020 (excel)".
18.6. Adjustment
Not applicable
18.6.1. Seasonal adjustment
Not applicable.
Annex 1: CVTS 2020 Questionnaire in English.
Annex 2: Instructions to interviewers on CVTS 2020 in English.
The Continuing Vocational Training Survey (CVTS) collects information on enterprises’ investment in the continuing vocational training of their staff. Continuing vocational training (CVT) refers to education or training measures or activities which are financed in total or at least partly by the enterprise (directly or indirectly). Part financing could include the use of work-time for the training activity as well as financing of training equipment.
Information available from the CVTS is grouped around the following topics:
- Provision of CVT courses and other forms of CVT (training/non-training enterprises)
- CVT strategies
- Participants in CVT courses
- Costs of CVT courses
- Time spent in CVT courses
- Characteristics of CVT courses
- Assessment of CVT activities
The CVTS also collects some information on initial vocational training (IVT).
There were no deviations from the definition of statistical unit as per manual. The enterprise definition is compliant with Council Regulation (EEC) No 696/93. More complex enterprise will be available from next year.
The enterprises covered were with 10 or more persons employed belonging to the certain NACE categories as per manual (see also 3.3).
The target population applied followed guidelines, with no deviations. It contains 2952 enterprises.
Variable A2tot (persons employed) refers to 31 December 2020.
No geographical deviations from Eurostat guidelines.
The reference year for CVTS 6 is the calendar year 2020.
A sample coverage of all local enterprises in operation within the employment and NACE strata defined in point 3.3 are provided by Business Register Unit within NSO. Any enterprises which are not operating or have less than 10 persons employed are excluded from the target population. Consistency checks are carried out with data provided from previous 5 years.
For CVTS 2020, the team managed to obtain most of the answers of the key and core variables of the questionnaire. The non-response of the other items were filled in by using hot deck imputation. Overall, item non-response of each variable was very low. No variable had an item non-response greater than 10%.
Simple random sampling is purposely used to ensure that the sample is uniformly distributed among different strata and to ensure that the population is accurately represented.
Number, EUR.
From the data entry programme, an extract of data was issued for basic data analysis following the rules established in the manual. Any queries regarding the received dataset was made at this stage, while additional imputations were tagged for later analysis. As already explained outlier detection was implemented and those companies which deemed to be outliers were given a weight of 1. The other weights in the same NACE and Employment size strata were adjusted accordingly to reflect the population. The steps in the weighting system were the following: first the initial weight was extracted, then another weight referred to as 'adjusted weight' was created. This 'adjusted weight' took into consideration the outlier detection procedure and each outlier was given a weight of 1 and other weights in the same strata were adjusted accordingly. The third and final weight disseminated is the calibrated weight which takes the 'adjusted weight' and calibrates it against the total employment of the population. In case of non-response of other variables (nor core or key), hot deck imputation was implemented so that no missing variables were transmitted.
See table 18.1 "Source data and data collection" in annex "MT - QR tables CVTS 2020 (excel)".
Every 5 years.
Calendar year refers to 2020. CVTS asks about the enterprise’s current status in the reference year 2020 while certain particular questions may also ask for information regarding the previous year.
No geographical comparability problems.
No additional variables related to COVID-19 were collected.
See table 15.1 "Comparability - geographical" in annex "MT - QR tables CVTS 2020 (excel)".
No additional changes were made between CVTS 6 and CVTS 5.
See table 15.2 "Comparability - over time" in annex "MT - QR tables CVTS 2020 (excel)".
