I noticed that for some observations there is a mismatch for the ethnicity variables between the files a_indall and a_indresp.
For example some individuals that are Indian according to the indresp file, result Pakistani according to the indall file.
I would like to understand the reason for the mismatch.
I know that the indresp file registers the interviewees'responses to the individual questionnaire, but what does the indall file register? Is it directly compiled by the interviewers?