National Survey for Wales
(replaces the Living in Wales Survey)


Main uses

The National Survey for Wales is a general household/adult survey for Wales.  As such it can potentially be used for a wide variety of analyses of households’ circumstances and adults’ views.  To date, however, there has only been one survey (for 2009-2010) and this was limited to being a pilot only.  As such, the number of analyses of the survey that can currently be conducted is limited.

The National Survey for Wales replaces the Living in Wales Survey, which was an annual survey conducted each year from 2004 to 2008.  In this context, the detailed material in the rest of this page relates to the Living in Wales Survey and not to the National Survey for Wales.

The National Survey for Wales and the Living in Wales Survey are the Welsh equivalent of the English Housing SurveyScottish Household Survey and Northern Irish Continuous Household Survey.


In summary:

  • Available from: UK data archive.
  • Registration required: yes.
  • First survey available: 2004.
  • Frequency: annual.
  • Updated: December.
  • Scope: Wales.
  • Format: SPSS.
  • Files: 4 files per year, with one at the household level and one at the individual level.
  • Documentation: none.
  • Weighted or unweighted: weighted.
  • Household income data: yes but gross only (not net), unequivalised and in ranges only.

General issues

Which software to use

As the annual dataset is around 19,000 records for individuals and 8,000 records for households, it can be exported into Excel.

When to use the individual and household datasets

For most purposes, the household dataset is the one to use as it contains the vast majority of the data, with very limited data in the individual file.

What weights to use

There is only one weight in the dataset, which is effectively a household weight.  For household-level analyses, this is clearly the weight to be used.  For individual-level analyses, however (and most of the questions in the survey concern the views of a single randomly chosen adult in each household), use of this weight risks over-weighting the results for single-adult households in comparison with households with two or more adults.  In this case, there are two choices: either use the household weight and present the results as proportions of households (effectively assuming that the views of the respondent fairly reflect the views of the household as a whole) or derive an individual adult weight by multiplying the household weight by the number of adults in the household.

What topics to analyse

Note that different questions are asked in different years.  Furthermore, some of the questions that were asked in all years have differing possible answers and are thus not strictly comparable.  The key is to look at the field name as well as its description: if the field names are the same then the questions are identical; if, however, the field names are different then the questions are in some way different even if the descriptions are the same.

Related to the above, some variables have two versions, one ‘du’ and the other ‘non-du’.   It is the ‘du’ ones which should be used as they apparently include handwritten answers.

Relevant graphs on this website

Welsh graphs

Indicator Table Graphs Comments
Dissatisfaction with local area household first Although this issue in principle relates to individuals, the data is actually in the household table.

Question not asked since 2005.

second Although this issue in principle relates to individuals, the data is actually in the household table.
Non-participation household all Although this issue in principle relates to individuals, the data is actually in the household table.

For those analyses which are by household income, allocate each household income group to a household income quintile, ensuring that the five income quintiles are of as similar a size as possible.