>> Now a lot of the data that researchers collect, survey data, are de-identified in the last segment when we talk about record linkage and data analysis, but also in the imputation set up. We'll point to ways on how to de-identify data, these techniques can be used for that very same purpose. And then the survey data or experimental data are put into data archives. This data archives is certainly worth visiting, and increasingly has administrative data in it as well. Likewise here, the Data Archives in the U.S., ICPSR is a very large one for the social sciences, super popular, has been around for years, and a similar one in Germany. Now, this is my vocation bias, providing you with these examples. I'm sure that you guys can for your own country, scoop out those datasets or data archives and share that with the community taking this course. What is interesting is that increasingly federal agency and national institutes that collect data, well actually even organizations like Federal Employment Agency in Germany that aren't part of the ones that collect the data as part of their administrative processes, make these data available for researchers in their own country or across the world. My current half employer, the Institute for Employment Research in Germany has put out these research data centers that allow accessing increasingly remote access to the data that are housed there. Obviously entirely de-identified, or with strong restriction on who can do that and how. But it is possible, and therefore, a rich resource for all of you when you do an analysis, and I can only recommend visiting these various websites. Data.gov is a website. If you haven't come across it you should definitely visit, and I'm aware that not all countries can access .gov websites. I apologize for those that can't, and I hope that will change for you in some form. But other countries have similar databases as well. This is a collection of government data available, and Eurostat similarly puts out a lot of data. Maybe not quite as many microdata that you can reanalyze for your own purposes, but certainly tables and aggregate figures that you can download as well. And the amount of stuff that's out there is updated all the time. What's nice about these kind of websites is they often pair the data up with message reports, so that you do really know enough about the data generating process that we talked about before. In general, there's a big movement to open data, open data forum, open data access. A lot of countries are increasingly subscribed to that because those running certain agencies or making policy decision, there are not enough people to analyze all these data. They rely on you to actually take this data and create good insights out of them. Data without boundaries is an interesting initiative that tried to, at least within Europe and beyond, to make this access of datasets easier in the exchange as well.