Processors dropped many variables in the NCS-R, NLAAS, and NSAL source datasets, for the following reasons:
To protect the identity of research subjects, including dropping source variables for constructed demographics created to mask identity;
If they were source variables for constructed variables created for other purposes (e.g., NSAL's variable E14a-e15a, which has combined data from three questions that resulted from routing respondents through three different question wordings to gather the same data); or
If they were the following types of computer-assisted instrument variables (the figure below shows examples of how these would appear in the NCS-R, NLAAS, and NSAL interactive codebooks):
Open-ended question with text response;
Open-ended "other specify" question with text response;
Instrument screen with introductory text ("ENTER 1 TO CONTINUE");
Interviewer checkpoint or instruction screen; or
Interviewing software routing check.
More information on dropped variables (including a list of the variables) can be found on pages 44-77 of the Data Processing Notes.