Skip to content

Accuracy and reliability

Contact info

Population and Education, Education
Jonas Holst-Jensen
+45 39 17 31 25

hoj@dst.dk

Get as PDF

Continuing Vocational Training Survey (CVTS)

CVTS2006: Postal questionnaires supplemented with data from administrative registers. About 60 per cent of the returned questionnaires ended up being scanned and the rest were registered manually. The scanning of the questionnaire gave problems with the data quality for some questions, in particular concerning the questions on amounts of money, number of persons and hours. CVTS2011 AND CVTS2016: Web interviews in combination with telephone interviews supplemented with data from administrative registers. The survey questions about amounts of money, number of persons and hours etc. in particular gave problems with the data quality as these questions needed several persons to be involved.

Overall accuracy

The questionnaire was sent to a selected sample of private enterprises.

For CVTS2006 there were 19.295 enterprises in the population and 2.819 enterprises in the selected sample. The response rate was 31.

For CVTS2011 there were 18.571 enterprises in the population and 3.036 enterprises in the selected sample. The response rate was 41.

For CVTS2016 there was 18.431 enterprises in the population and 3.001 enterprises in the selected sample. The response rate was 50.

Measurements have been taken to compensate for the item non response. The information from the enterprises not answering all quantitative questions has been imputated. The enterprises are divided to 60 categories by economic activity (20 NACE categories) and number of employed persons (three size categories).The imputation replaces missing data with data from enterprises with identical characteristics concerning the size and the economic activity of the enterprise. Measures are also taken to reduce unit non-response based on a weight applied to each data record. The method of weighting is based on information on number of persons employed from the register of enterprises and carried out by a regression estimation.

Sampling error

CVTS2016. The uncertainty as expressed by the variation coefficient for the one of the main variables: Enterprises offering CVT: The variation coefficient tells the uncertainty as a percentage of the mean of the variable. Based on the estimated value of the variable (i.e. the value related to the universe) a 95 percent interval of confidence can be expressed as plus/minus twice the variation coefficient.

For the total sample (I.e. All) the uncertainty is plus/minus 5,23

The uncertainty is larger for the small and large sized enterprises than for the middle size. For the small enterprises the uncertainty is plus/minus 3,11 and for the larger it is plus/minus 4,86. For the NACE-groups the largest uncertainty is for group 1: plus/minus 16,43.

All: 5,23

Size groups:

  • 1: 10-49 empl.: 7,96
  • 2: 50-249 empl.: 9,22
  • 3: 250+ empl.: 6,86

Branches:

  • Nace 1: 29,27
  • Nace 2: 14,10
  • Nace 3: 17,46
  • Nace 4: 10,86
  • Nace 5: 11,15
  • Nace 6: 10,86
  • Nace 7: 10,16
  • Nace 8: 13,02
  • Nace 9: 10,17
  • Nace 10: 18,48
  • Nace 11: 10,81
  • Nace 12: 11,72
  • Nace 13: 25,97
  • Nace 14: 13,93
  • Nace 15: 17,64
  • Nace 16: 19,49
  • Nace 17: 20,27
  • Nace 18: 15,61
  • Nace 19: 50,97
  • Nace 20: 24,47

CVTS2011: The variation coefficient for enterprises offering CVT:

All: 2,52

Size groups:

  • 1: 10-49 empl.: 3,11
  • 2: 50-249 empl.: 2,68
  • 3: 250+ empl.: 4,86

Groupings of enterprises (NACE-groups):

  • Nace 1: 16,43
  • Nace 2: 6,42
  • Nace 3: 7,02
  • Nace 4: 6,78
  • Nace 5: 7,69
  • Nace 6: 7,45
  • Nace 7: 6,35
  • Nace 8: 5,18
  • Nace 9: 7,75
  • Nace 10: 5,43
  • Nace 11: 9,96
  • Nace 12: 7,43
  • Nace 13: 6,99
  • Nace 14: 7,87
  • Nace 15: 8,29
  • Nace 16: 8,86
  • Nace 17: 6,57
  • Nace 18: 6,72
  • Nace 19: 9,66
  • Nace 20: 6,53

Non-sampling error

The enterprises are sampled based on their CVR-number. However, the activities of some companies are involving several CVR-numbers which means that the answers in the interview can't be related to the specific CVR-number sampled. As various people in the enterprise had to answer specific parts of the interview, a number on inconsistencies in the answers was found in the data. The follow up and corrections necessary demanded a considerably effort.

Quality management

Statistics Denmark follows the recommendations on organisation and management of quality given in the Code of Practice for European Statistics (CoP) and the implementation guidelines given in the Quality Assurance Framework of the European Statistical System (QAF). A Working Group on Quality and a central quality assurance function have been established to continuously carry through control of products and processes.

Quality assurance

Statistics Denmark follows the principles in the Code of Practice for European Statistics (CoP) and uses the Quality Assurance Framework of the European Statistical System (QAF) for the implementation of the principles. This involves continuous decentralized and central control of products and processes based on documentation following international standards. The central quality assurance function reports to the Working Group on Quality. Reports include suggestions for improvement that are assessed, decided and subsequently implemented.

Quality assessment

The questionnaire was sent to a selected sample of enterprises. To reach the final result the enterprises were weighted in relation to their size (by number of employees) and financial activity (by NACE codes).

For CVTS2006 there was 19.295 enterprises in the population and 2.819 enterprises in the selected sample. The response rate was 31.

For CVTS2011 there was 18.571 enterprises in the population and 3.036 enterprises in the selected sample. The response rate was 41.

For CVTS2016 there was 18.431 enterprises in the population and 3.001 enterprises in the selected sample. The response rate was 50.

For CVTS2011 and CVTS2016 Some of the enterprises who had answered the questionnaire had not answered some of the questions. In the case of quantitative data the information was imputed (core variables). The imputation substitutes the missing data from the enterprises with the same characteristics. The criteria used was financial activity and size (number of employees).

Many working hours were executed in the process of ensuring logic and a reasonably consistency in the answers given by the enterprises. E.g. the relations between people who had been on courses, the number of working hours they had spent on this as well as expenses were not always logic and consistent.

Data revision - policy

Statistics Denmark revises published figures in accordance with the Revision Policy for Statistics Denmark. The common procedures and principles of the Revision Policy are for some statistics supplemented by a specific revision practice.

Data revision practice

Only final figures are published.