So I am working on a research project where I need to combine two datasets from a community survey given every year in NYC. I am combining years 2016 and 2017 which have the same variables used for each year, and each year has a respective survey weight. I am using the append function to combine the datasets but I am unsure how to combine the survey weights for a total survey weight of the two. I'm working with Stata 14.2 if that helps at all. I'm new to stata so if you can clearly explain what you have to do/why you do it, that would be really helpful. Any insight on this would be great! Thanks!
Asked
Active
Viewed 654 times
1 Answers
0
Assuming that each file represents the same population (all NYC or whatever), divide the survey weight on each observation by 2. Just remember what population you are now making inferences for. You now have an average population over the time period 2016-2017.
In fact, you don't even have to divide by 2, but doing so will save you some explaining later.
If the two surveys represent two completely different populations, the answer is probably don't do anything, but I'm not sure who you are making inferences for.
If the two surveys represent some (but not all) of the same population, take the overlap portion and divide those weights by 2.
