Statistical processing

Source data

The survey is based on questionnaires. The source for the sample and updating of company specific information is the Central Business Register. Annual investment information from Regnskabsstatistik (Industrial accounts statistics) forms the basis of the weighting system.

The survey is carried out with a panel based sample consisting initially of 800 enterprises representing about 65 pct. of the total number of full-time employees in selected trades within the included strata.

Because of drop-out the sample is updated ones a year. Typically the sample is added 100-120 new enterprises at update.

Frequency of data collection

Data is collected twice a year (March and October).

Data collection

Data is collected either by paper questionnaire send to the respondent including postage pre-paid envelope, or by online reporting. The cover letter urges to log in to the reporting system VIRK, and fill the online questionnaire.

Reminder is send about the 25th in the survey month, with new deadline about the 4th next month.

Response rate is 80-85 pct. Deadline for responding is about the 20th of the month. Here, about two third of the answers have arrived. The rest, one third, that arrives within four weeks and are included in the statistics.

Data validation

Data validation of returned questionnaires is done manually. Figures are compared to previous reported figures. Also is compared to latest available investment figures from the annual accounts statistics.

The filled paper questionnaires are scanned automatically. If rejected the automatic scan they are handled by manual entry. By on-line reporting the system warns the respondent if a question is not ticked, however some issues can be missed on purpose from the reporter's hand and validation is done manually as described above.

Data compilation

As far as possible lacking response figures are imputed from previous responses from same enterprise. Normally the imputation rate is about 9 pct. (un-weighted).

If the sample has been updated with new respondents, historical data is lacking for the year-by-year change calculations. Those data are impute them from the Accounting Statistics. Each figure is evaluated if it looks realistic (different workplaces and industries)

In the enumeration process the sample is clustered in 7 record groups. Within each record group is calculated the growth rates of the units that have responded in the relevant periods.

Total investment figures for the year T is based on the latest Account Statistics (+ 18 month) from year T-2. So, first is calculated the total investment figures for year T-1 by multiplying with the growth rate figures for that year. In second step those figures are multiplied with the present growth rate figures, and absolute figures in DKK for the total investment is the result.

the 7 record groups refers to the EU's classificationsystem [Main Industris Grouping (MIG)[(https://ec.europa.eu/eurostat/statistics-explained/index.php/Glossary:Main_industrial_grouping_(MIG)).

• 29 Manufacture of motor vehicles, trailers and semi-trailers
• CDUR Durable consumer goods
• CNDU Non-durable consumer goods
• FOBE Food and beverages industry (10.1-10.5, 10.7-10.8, 11, 12)
• CONS Consumer goods (CDUR+CNDU)
• INTM Intermediate goods (excluding mining and quarrying)
• INVE Investment goods (includes 29)

Out of those Statistics Denmark dessiminates four groups: INVE; INTM; CDUR and CNDU along with the total.

Adjustment

No adjustments are carried out.