chapter 3 discussed data biases resulting from nonrandom patterns in report filing and detection, as well as peculiarities of the registration and categorization process more generally. The current chapter examines the quality and validity of the study data, and provides a description of the variables used.

The review of the sample on which the study is based also includes a description of data selection, data supervision, and of reformatting and other changes the data have undergone prior to analysis. In addition to documenting the quality of the data and the processing it has undergone, this review provides an overview of the production of the crime statistics.

The Primary Sample

The advantage of a study based on electronic data is that larger samples are just as easily studied as smaller ones. Since register research typically allows for the use of data from nationwide registers, very large samples become easily analyzable. Such is the case with the electronic data on which this study is based.

As crime is a relatively rare phenomenon, a large sample has been defined so that it is possible to perform a number of special analyses. Since some of the analyses require comparison with the population in general, the primary sample has been drawn from a register that includes all residents in Denmark. “Residents” are individuals with a valid Danish CPR number (Central Person Registration Number).

The primary sample has been drawn to include one-fifteenth of those with a valid CPR number.