- How are missing values treated in the data?
- How do I identify imputed values?
- What are derived variables?
Missing Values: Analysts should take care because no system missing values are set in the BHPS data released to the UK Data Archive. Missing data, either due to refusal, inapplicability or otherwise are coded -1 to -9 but are not otherwise set to system missing status for use in any statistical analysis package. See Missing Value Conventions, in the user documentation for details.
Derived and Imputed Variables: In addition to the actual answers to the survey questions, BHPS staff derive certain variables from data provided by respondents including household income, various socio-economic class indicators, etc.,… Missing values for a range of income and other variables are also imputed by BHPS staff. Flag variables are provided to distinguish the imputed cases for a given variable. Please see the Notes on Derived Variables and Data Imputation sections in the user documentation for more details.
In addition to derived variables, certain additional data sets comprised of derived data from the BHPS are obtainable from the UK Data Archive. Included here are the “Net Income Files” and “Employment History Variables” derived from retrospective employment histories and panel data. Please see the BHPS titles available through the UK Data Archive web-site for details.