Colleague network (from 01-01-2009)

Description

This network layer contains all peer relationships of individuals who are registered in the Personal Records Database (BRP) on 1 January YYYY and who are employed in the Netherlands. Colleagues are considered to be persons who are registered in the BRP on 1 January YYYY and who are employed by the same company as the person in question. The basis of this stock are the (S)POLISYYYYBUSVV, the (S)POLISLONGBAANTABVV and the GBAADRESOBJECTYYYBUSVV. This file is part of a network file set. This network file set also contains the following files: BURENNETWORKYYYYTABVV, FAMILYNETWORKYYYYTABVV, HOUSEGENOTENNETWORKYYYYTABVV and CLASSGENOTENNETWORKYYYTABVV. These files have the same structure. The colleague network file has been compiled as well as possible on the basis of logical assumptions and available data. It contains assumptions, artifacts, and noise. They are suitable for many research purposes, but not for statistics or counts of numbers of relationships. For example, you can't make statistics about how many co-workers someone has, because co-workers are artificially determined and the number of co-workers is capped (see below). For some research questions, the limitations of the files will be more relevant than for others. Much of the noise is mediated on a large (research) scale. But in all cases it is important that researchers who work with the data take due account of the underlying derivation rules, and if in doubt consult a CBS expert in the field of these networks. More information on how to access the data: > https://www.cbs.nl/en-en/our-services/custom-and-microdata/microdata-self-research ### Methodology For each person, a maximum of 100 colleagues are derived. This involves looking at the people who work at the same company ((S)BEID). For companies with fewer than 100 employees, this means that all persons in that company have colleague relationships with each other. For companies with more than 100 employees, the 100 closest living colleagues are selected. The distraction of the colleague consists of two steps. In the first step, it is checked per company per person per month who the maximum of 100 closest living colleagues are. After all, colleagues can leave, someone can change jobs themselves, and someone can have multiple jobs at different companies. So per month and company where someone works, someone has a set of up to 100 colleagues. In the second step, those sets per person are merged into one large set, from which again the maximum 100 closest living colleagues are selected.In this way, regardless of the number of companies in which someone works during the year, everyone has a maximum of 100 colleagues. Because there is no information available about branches and departments within companies, it is not possible to infer which people actually have contact during their work. This problem may be partly overcome by selecting the colleague based on the proximity of their home address. In addition, colleague's are derived based on the company that pays the employee's employment compensation. It may therefore happen that the identified colleague network is less in line with the actual colleague network for some persons, for example for persons who are posted or work as a temporary worker. It can therefore not be said that the colleague 's that are derived in this file actually have contact with each other. They can be seen as a reservoir of potential contacts. It is possible that people have multiple jobs at the same time. When these are multiple jobs within the same company ((S)BEID), only one of these jobs is included in the first step of the distraction of the colleague. If the jobs are at different companies, then all jobs are included in the first step. Because many companies have more than two employees, in the very large majority of cases there will be more than one colleague relationship per RINPERSON. The network is not symmetrical. This means that person B can appear in the co-worker network from person A, but person A does not need to appear in person B's co-worker network. This will occur in individuals working in companies with more than 100 employees, due to the selection of co-worker based on distance. ### Population This network layer contains all peer relationships of individuals who are registered in the BRP on 1 January YYYY and are employed in the Netherlands. Colleagues are considered to be persons who are registered in the BRP on 1 January YYYY and who are employed by the same company as the person in question. This file is not suitable to reproduce official population numbers on 1 January according to Statline. For this purpose, the component GBASTANDPOPULATIONYYYYTABVV must be used.

Resources

Name Format Description Link

Tags

  • werkkring
  • persoonsnetwerk
  • werknemer
  • microdata
  • data-governance-act
  • odissei
  • dga

Topics

Categories