AHTUS Missing State IDs


Hi, I downloaded an AHTUS extract with all samples from 1993 to 2012. There are over 3 million observations, but only about 150,000 have state information. Why is this?


It sounds like you are working with a hierarchical extract, which would be why the state information appears to be missing. In this format, the data contains person records followed by the activity records. As such, you will only have one observation (row) per person containing information such as state, age, sex, etc. We have a bit of information on rectangular versus hierarchical data here.

If you revise your extract to be rectangular, you will see that you have about 150,000 observations total, all with state information.