What does the survey comprise?
The survey ‘ICT use by companies’ collects annual data on automation and the use of information and communication technology (ICT) in companies in the Netherlands. Among other things it describes the use of computers, the internet, electronic buying and selling, software and ICT applications. The results of this survey are important for policymakers, researchers, companies, and trade and industry associations.
Since 2003, Statistics Netherlands has conducted the survey on ICT use by companies among companies based in the Netherlands with 10 or more employed persons. The population contains about 60 thousand companies. In 2002, the threshold was still 5 employed persons.
Statistics Netherlands maintains a register of companies that can be used as statistical units. These are the actual performers in the production process. They are characterised by autonomy regarding decisions about that process and by offering their products to third parties.
Date/year survey started
The study was started in 1983 more or less in its present form.
The survey is conducted once a year.
The first figures published are definite. No data become available later that could lead to any further adjustments. Results are not adjusted as a result of questionnaires received late.
The results of the survey ' ICT use by companies' are published in StatLine tables on Statistics Netherlands’ website. The results are also the basis for the series ‘The digital economy’, which has been published annually by Statistics Netherlands since 2001. Every year, this book comprises numerous data on ICT use in the Netherlands and in an international perspective.
How is the survey conducted?
The annual survey 'ICT use by companies' includes a random sample of the companies in the target population. The survey is conducted both via internet as with paper questionnaires. The reference date is December 31.
Companies based in the Netherlands with 10 or more employees.
The sample consists of about 11 thousand companies annually.
Checking and correction methods
The data are first made consistent at the individual level by correcting conflicting answers. Later, the aggregated data are compared with the figures for the previous year, where large discrepancies are checked again and where possible corrected.
The companies are classified by a combination of business activity and size (cell). The results are based on a stratified sample: the results for each cell are weighted by the ratio of the number of companies per cell in the population (N) and the number of companies per cell in the net sample (n).
What is the quality of the results?
This survey is conducted among a representative sample. Outcomes are counted before and after processing to detect errors. Item non-response is only estimated for quantitative questions.
The population framework will always contain errors, namely under-coverage, over-coverage and classification errors. Over-coverage is a result of units wrongly included in the framework, e.g. those that no longer exist or have been taken over. Under-coverage is a result of companies not included in the framework as they are not yet registered. Classification errors arise when a company is included in an in incorrect business sector or because they are assigned a wrong company size.
Comparing results of this survey in time presents some problems, as the structure and sequence of questions in the survey change annually. In addition, questions are frequently reformulated to match new insights. These changes in the questionnaire may cause breaks in time series.
The data are first made consistent at the individual level. The aggregated data are then compared with the figures for the previous year. Furthermore the distribution margins are calculated for the main variables.