With micro data from different countries it is usually the case that the survey in each different country is a different size, and the different sample sizes may have nothing to do with the number of people in each country. When analysing this kind of cross-national data you usually interested in either saying something about the group of countries as a whole, or in treating each country as having equal weight in comparison with others. Both situations require you to weight the data for analysis, but in different ways.

Another complicating factor is that there are sometimes survey sampling weights for the micro data that account for either the structure of the sampling scheme or the pattern of unit non-response, or both. For what follows you can assume that these weights are in a variable called surveyweight. If there are no such weights, you can set this variable equal to 1 and the following commands will still work.