Frequently Asked Questions
Below is a list of frequently asked questions (FAQ), curated by the POPNET team. If you have additional questions or comments, please contact us via email.
Where can I download the data you work with in the POPNET project?
The social network data of the Dutch population is currently not publicly available. One of the main goals of POPNET is to unlock (aggregated) (parts of) the data, strictly for academic research. But we need to do this in a secure way, and make sure that for instance anonymity is guaranteed. We also need a proper platform in terms of hardware and software where users can analyze the data. Researchers at Dutch universities can apply for secure access to a simplified version of the underlying raw data at Statistics Netherlands, see CBS Microdata.
Who is the owner of this data?
The social network data is compiled by Statistics Netherlands / Centraal Bureau voor de Statistiek (CBS). CBS has the legal mandate to collect this data from various governmental agencies, and use it for statistical research. From this data, CBS compiled the social network data. The data is currently stored at secure internal servers of CBS.
I am a researcher and this data is great for a project I am working on. Can we work together?
Good to hear! We see a wide range of research possibilities with this data, and that is the reason we work on this research infrastructure. Feel free to send us an email with your idea. However, currently the data is only available in the secure non-publicly accessible server environment of Statistics Netherlands. When we are further in the project, we hope to be able to responsibly share (aggregated) (parts of) the data for research purposes. Knowing your research interests and questions will help us to decide on how and what data we will make available.
Can you find information about specific persons in the data such as myself, or my parents?
No, we cannot. The data is pseudonymized: there are no names or personal identifiers such as your Burger Service Nummer (BSN) in the data, so we do not know who is who. The data also does not contain personal information such as address or date of birth.
Are you able to find particular people in your data based on their very different profiles?
Good question. In fact, in rare cases it might indeed be possible to locate a person in the data, based on a particular unique pattern in the connectivity of a person’s social network. Of course this requires first of all full knowledge about this person’s social connections. To reduce these risks, and in addition to pseudonymization, we develop and employ algorithms that measure these risks as precisely as for tabular data (as is done in traditional statistical disclosure control), so that any (aggregated) data that in the future will become available through the platform, remains properly anonymized.
I am registered in the Netherlands but do not want to be in your dataset. What kind of data do you have about me? Can you remove me?
We cannot find you, or remove you, since our dataset is pseudonymized. We do not have access to social security (BSN) numbers, names, addresses, or dates of birth. You can to some extent request to revoke access to your data for particular (research) purposes, through your municipality. See Rijksoverheid information on BRP.
You do academic research. Is your research output always available?
Yes, papers that we write, or presentations that we give, are published in open source journals, or presented at conferences that typically anyone can attend.
Do you work with commercial parties or intelligence agencies?
No, we do not.
Do you really need information about the entire population for your research? Why can’t you just use a sample?
Much academic work can be accomplished with samples. But we also know that samples are unable to provide enough input for dynamical network processes such as the modeling of epidemic spreading, for example. Therefore, one of our aims is to show how population-scale network information can lead to better insights and different results from what we could obtain through surveys such as the LISS panel.
Which methods were used to derive the network?
This CBS working paper describes the approach used to derive the network. Stay tuned for papers by our team on the structure of this data!