For variables asad, anervous, ahopeless… and the other mental health indicators used to generate the K6 dperession indicator, there seems to be a response rate of about 68% (years 1999-2010), and 32% NIU. Was this because it was only asked 2/3 of the time? Also, I checked the variables and they are all the same missing across the variables.
For asad, anervous, ahopeless... why so many missing variables? Was this asked only 2/3 of the time?
Only sample adults were asked the mental health questions. A sample adult is the one adult (age 18+) per family who was selected at random to answer additional health-related questions, including the mental health questions. Thus, any person that does not have ASTATFLG=“1:Sample adult, has record” are coded as NIU for all of the mental health questions.
Keep in mind that analyses using at least one variable that was answered by only sample adults should use SAMPWEIGHT instead of PERWEIGHT when weighting the data.
Hope this helps.