Understanding Society User Support

Is there an equivalent in wave 2 for the variable a_swemwbs?

No, the Warwick-Edinburgh Mental Well-being Scale (http://www.healthscotland.com/documents/1467.aspx) has so far only been carried at Waves 1 and 4. See content plan here: https://www.understandingsociety.ac.uk/d/18/Long_Term_Content_Plan_Nov2011-3.pdf?1355920157

Can you recommend a variable which could be use as a substitute for deprivation (since index of deprivation is not available)?

Small area indicators including measures of area deprivation from external sources can be linked to UKHLS data. Please see links to the application process here: https://www.understandingsociety.ac.uk/support/projects/support/wiki/Is_it_possible_to_merge_in_data_on_geographical_identifiers_eg_Census_information

Many different measures of socio-economic status could be used or constructed from UKHLS data. The study collects detailed occupation and income information and several derived variables for occupation-based classification, personal and household incomes are supplied with the dataset. See user guide for further information: https://www.understandingsociety.ac.uk/documentation/mainstage/user-guide or our publications on e.g. poverty may also be of interest: https://www.iser.essex.ac.uk/search?q=poverty

There is a large number of missing values in the variables b_sf12pcs_dv and b_sf12mcs_dv even when using filters for full interviews and adult self-completion only (i.e. using select cases for b_ivfio_dv=1 and B_scflag_dv=1). Can you advise me if these are real missing values or if there's an exclusion I'm not aware of? should I be using any other filter if I'm interested in individual responses for these variables?

A few of the underlying SF12 items have a higher level of non-response than others. This is not unusual for self-completion questionnaires where participants may skip questions they don’t feel they apply specifically to their circumstances. The composite variables have only been derived for complete cases. Analyst may here choose to use the derived variables supplied with the dataset or impute the missing items and construct their own composite variables.

Jakob