Hello, I'd like to get anyone's thoughts about what seems to be a data artifact but I am hoping for further clarification or to know if I might be missing something. I'm interested in past 30 day alcohol use among currently pregnant women and I get the following weighted percentages using the raw PUF data:
Wave 1 - 4.7%
Wave 2 - 10.2%
Wave 3 - 9.3%
I used the cross-sectional weights for wave 1, the single-wave weights for wave 2, and the all-waves weights for wave 3.
The 4.7% estimate for wave 1 seems to be a little low. I wouldn't expect it to be half of the other two and I'm wondering if anyone has any ideas??