Development

The National Survey for Wales was given approval by the Welsh Cabinet in the summer of 2010.  Since then, numerous development work has taken place to ensure the survey would be fully effective when the main stage of fieldwork was launched (January 2012).Work to ensure the survey design continues to offer value for money and takes advantage of the latest social research and technical developments are on going.

An extensive programme of cognitive research took place to develop the survey materials and to ensure the questions used in the survey were being understood by respondents as they were intended. Both the survey materials and the questions were further tested during a pilot test and field test carried out in late 2011. The results of these tests helped to iron out any remaining issues with the survey before the main stage of fieldwork.

Research was commissioned to investigate whether the collection of interviewer observation variables for both responding and non-responding households could increase the accuracy of National Survey for Wales estimates through non-response weighting.

Further information on this work can be found in the detailed reports in the left hand menu.

Work recently completed

The Welsh Government has also recently commissioned research into three methodological issues facing the National Survey for Wales. One report looks at the issues we face when generating aggregate statistics, another investigates when it's appropriate to compare results from questions in the National Survey with other surveys and the third report looks at the possibility of mixing survey modes in future. These reports can be found under ‘survey methodology reports’ in the left hand menu.

Upcoming work

Further non-response development work is planned for later this year. Three new interviewer observation variables have been implemented into the 2012-2013 questionnaire; the new variables have a stronger association with key variables in the questionnaire. It is hoped that work will be commissioned to analyse the effects of all of the observational variables in the survey to identify recommendations for the optimal combination of variables to be used in future rounds of the survey to reduce non-response bias.

A piece of work is being carried out to look at what influences Welsh speakers to participate in past National Survey for Wales studies in English rather than Welsh. The results of this research will provide some insight into participation through the Welsh language and give some recommendations for measures that will help increase the proportion of those respondents who complete their interview in Welsh, which currently stands at 2% on average. It is hoped this work will be completed within the next couple of months.