The Wave 1 panel consisted of around 5,500 households and 10,300 individuals drawn from 250 areas of Great Britain. Additional samples of 1,500 households in each of Scotland and Wales were added to the main sample in 1999, and in 2001 a sample of 2,000 households was added in Northern Ireland, making the panel suitable for UK-wide research. As this was a longitudinal household study, households were sampled and then individuals living in these households were followed wherever they moved to within the UK. All 16+ year old household members were eligible for adult interviews, while from Wave 4 onwards, all 11-15 year olds were eligible for youth self-completion questionnaires. Interviews were conducted face-to-face with a few by proxy interviews and a few by telephone.
As part of wave 18, BHPS participants were asked if they would consider joining the new, larger and more wide-ranging survey Understanding Society. Almost 6,700 of just over 8,000 BHPS participants invited to join did so. The continuing sample from the British Household Panel Survey (BHPS) joined the Understanding Society sample in Wave 2. The cases in the two samples can be distinguished using the variable w_hhorig. The variable also allows the identification of different components of the BHPS sample (see below). This will allow continued research on changes in people’s lives from 1991.
The sample design, data structure, interview pattern, following rules are very similar across the two surveys. The core questions used in BHPS were included in the new study to allow researchers to study trends and changes in people’s lives starting from 1991. There are, however, a few differences in the data collected. New topic areas were included such as more questions related to health and wellbeing, ethnicity and migration, social networks, and topical issues such as the Olympics and EU referendum.
Harmonised or stand-alone BHPS
To use the long-run of data collected from BHPS sample members along with the data collected as part of Understanding Society, users are advised to use the harmonised BHPS, which is included in the Understanding Society data release. If using the stand-alone BHPS (SN5151) then they will have to undertake the harmonisation exercise themselves.
Both options facilitate linking cases across studies using the unique UKHLS person identifier pidp which has been added for all BHPS sample members in the BHPS data files (i.e., irrespective of whether they have ever participated in the UKHLS).
There are advantages in using the harmonised BHPS files: all variables that are equivalent in both studies have been renamed so they have the UKHLS name and efforts have been made to assure that the information content is identical. If the name of a variable was the same but the content was not identical then the BHPS variable received the suffix _bh. All BHPS variables with names that do not correspond to any variable in UKHLS are left as is.
The same convention was followed for data files. The xwavedat files from both studies have been merged and all information that is available for cases from both samples has been harmonised. In this harmonised version, all BHPS Wave 1-18 file names and variable names have a prefix “bw_”.
Harmonised BHPS data are documented in the Understanding Society online data documentation, including the variable occurrence stretching across both studies removing the absolute requirement to jump across study documentations.
Users should note that the harmonisation project is ongoing and a number of data aspects that could be harmonised in principle have not yet been harmonised due to the complexity of the task and time constraints. More detail about the harmonised BHPS is provided in the designated Understanding Society harmonised BHPS User Guide.
If using the stand-alone, non-harmonised, BHPS data (SN5151) it is important to be aware that variable names in the BHPS dataset have slightly different formats:
- They are limited to eight characters.
- There is no underscore separating the wave prefix from the main part of the name, so indresp file for Wave 1 is named aindrep and sex variable for Wave 1 is named asex.
- Derived variables, imputation flags, weights and other special variables are not distinguished by “_dv” or “_if” suffixes.
- As this is the stand-alone non-harmonised version, variables with same names as in the UKHLS files may not have the same meaning.
