How is the survey comprised?
The Netherlands’ Safety Monitor (Veiligheidsmonitor-VM) monitors the attitudes and experiences of the general population of the Netherlands regarding habitability and safety of the neighbourhood, fear and prevention of crime, victimisation and the attitudes towards the police and municipality.
It provides information down to a very detailed regional level (for some municipalities even to the level of neighbourhoods).
Persons aged 15 years or older living in private households.
Persons, households, delicts, contacts.
Year survey started
A separate safety survey has been conducted since 2005* in the Netherlands. The questionnaire and the design of the survey were adjusted in 2008 and in 2012. The survey has remained unchanged since 2012.
The VM is commissioned by Statistics Netherlands, the Ministry of Security and Justice and regional and local (government) agencies.
*Before 2005, safety was included in other general surveys.
Annually. The survey takes place in the period August until November.
The figures are definite. Until now only published in Dutch.
How is the survey conducted?
The VM survey is a sample survey. The sample is taken from all non-institutionalised Dutch residents who are 15 years or older and are registered with their local municipality. From this group a stratified sample is taken according to the design of the survey, with nationwide coverage of municipalities.
The VM is conducted according to a ‘mixed mode’ design. Sample units (persons) are invited by letter to participate via the internet (CAWI – Computer Assisted Web Interviewing). Non-respondents are re-approached twice, each time a paper questionnaire is included. Persons who still don’t respond on this are reminded by telephone if their telephone number is known.
Persons of 15 years and older living in private households.
Number of respondents
The number of respondents differs every year. There is a nationwide minimum number of 65 thousand respondents to obtain reliable figures on police district level.
In addition, a different number of municipalities participate each year to obtain information for their neighbourhoods. As a result, the total number of respondents ranges from 80 thousand to 145 thousand in the period 2012-2016.
Checking and correction methods
Internal consistency and completeness is checked. With the online questionnaires this happens during the interview. Paper questionnaires are read automatically during which inconsistent and incomplete answers are adapted with specially designed control and correction programs.
The raw data are weighted to compensate for differences in personal and regional characteristics between the sample and the population. The weighted results are scaled-up to represent the entire population per region.
Per region*, the weighting factor adjusts for differences in sex, age, ethnic background, household size and household income.
*Region= police district, province, municipality or neighbourhood
As the survey is based on a sample, the results will fluctuate because of random errors. The accuracy of the estimates is indicated by the ‘95% confidence interval’, which contains the true (unknown) value with 95% certainty. The confidence intervals are calculated using the SPSS Complex Sample Module to allow for the stratified design of the survey. Figures are not published when a point estimate is based on fewer than 50 observations.
The data are comparable over the years since 2012. Comparable figures have been obtained for several indicators between 2005 and 2012 by applying corrections. These corrections have been calculated to adjust for differences in methodology and questioning.