This survey are sample surveys, where a representative sample of persons 16-74 years are asked among other things about the consumer expectations. The results are corrected from the effects of non-sampling and non-response and then enumerated so that the figures can directly be classed with the population of adult persons and families in Denmark. Data are validated using logical validation rules. A seasonal pattern could not be identified in the series and no seasonal adjustment was undertaken.
The data is collected through Statistics Denmark's omnibus surveys. The information used for the purpose of the consumer expectations survey has always been collected as part of Statistics Denmark's omnibus surveys. These surveys are sample surveys, where a representative sample of persons 16-74 years are asked among other things about the consumer expectations. The results are corrected from the effects of non-sampling and non-response and then enumerated so that the figures can directly be classed with the population of adult persons and families in Denmark. The data collection takes place within the first two weeks of the month, as well as a few days before the first day of the reference month in order to achieve a satisfactory response rate.
An omnibus survey.
In the frame of data validation the check of the data correctness is conducted by checking the consistency of the responses by every single respondent, comparison of the results with the results of the previous month, as well as comparison of the non-response rate are also undertaken. There is also a built in formulae in Excel to check whether the results comply with the output from SAS program. Otherwise data are validated using logical validation rules.
As it is not possible to interview around 34 percentage of 1500 persons, each completed interview is given a personal weight, which partly corrects the effects of non-sampling and non-response and partly ensures that if you sum up all completed interviews you end up with a number of persons in tables and analyses corresponding to a number of persons of the age between 16 and 74 years. The questions asked to the persons are being enumerated with this weight, as well as weighted figures are being used by calculations.
However, a number of variables do not concern the individual person, but the family he/she is living with.
In these cases each respondent who is answering is being given a family weight, which ensures that the families with different numbers of persons aged 16 years or over are represented in the sample with varying probability. A family is defined as one or several persons between 16 and 74 years who live in the same accommodation and are related (includes also cohabiting couples) and their children, though there can only be one married or cohabiting couple in the family. The sum of the family weight for all completed interviews, gives the number of families in Denmark that have persons aged between 16 and 74 years. The number of families is thus an estimate for the random test.
No furtherer adjustments are undertaken, beyond those that have already been described under Data validation and Data compilation.