Understanding Society User Support

Hi there,

I am working on waves 1, 3, and 5 of Understanding Society. I want to use the wave 1 "jbnssec5_dv" variable as an indicator of socioeconomic status, however, I have identified some potential issues with this variable that I cannot currently understand. The response options for this variable are: management & professional, intermediate, small employers & own account, lower supervisory & own account, lower supervisory & technical and semi-routine & routine. However, this variable has a high proportion of non-applicables at wave 1. I assumed these non-applicables were likely due to respondents indicating that they were not in paid employment, which did account for a good proportion of the non-applicables. However, to check that this was the main reason for NA responses, I also checked if any respondents indicated they were not in paid employment but had viable responses to jbnssec5_dv (i.e., management & professional, intermediate, small employers & own account, lower supervisory & own account, lower supervisory & technical and semi-routine & routine), which I identified (in my subsample of respondents) that over 8,000 respondents indicate no paid employment but then state they're employment is either: management & professional, intermediate, small employers & own account, lower supervisory & own account, lower supervisory & technical and semi-routine & routine. Should respondents who are not employed have viable responses to this variable or should they be coded as NA?

Many Thanks