How can we continue to accelerate the creation of robust and transparent research that allows us to better understand – and improve – people’s wellbeing?
Launched today, our new wellbeing data usage library aims to increase the accessibility of wellbeing data for researchers.
Interim Head of Evidence at What Works Centre for Wellbeing and Professor of Public Policy at KCL, Michael Sanders, explains what we are doing and why.
For over a decade, Britain has made strides forward in capturing wellbeing across a number of datasets, including the Civil Service People survey, Annual Population Survey, and others. Together these datasets allow us to better understand people’s wellbeing and how it changes over time.
We are now able to use robust consistent wellbeing evidence and measures with confidence. And, there is still more to learn about how to improve wellbeing, and how wellbeing levels change for and affect different people.
To build more knowledge about wellbeing, we need more researchers to get involved in using existing wellbeing data. We can do this by making it more accessible and easier to use by reducing the ‘cost of entry’.
Our wellbeing data usage library
We have published free-to-use code for data cleaning, hosted on the collaborative software development platform GitHub. Our page hosts repositories containing code we have used to extract and clean wellbeing data from specific datasets for a number of projects.
The benefit of this initiative is that, rather than spending days familiarising yourself with the data and cleaning it to make it usable, the code can reduce this to hours, saving valuable working time. This builds capacity and capability.
How has the Centre used the code?
Our Github page currently holds repositories with code for three of our projects:
UK time use data records the enjoyment associated with people’s activity. We are collaborating with University College London’s Centre for Time Use Research to explore how people are living their lives, how they respond to external changes, and how this impacts their own wellbeing and that of the wider society. The value of this data lies in the potential to produce new policy-directed insights to enhance wellbeing.
For this work, we used three Stata Do files to prepare the data for analysis:
- Using the data and creating new variables
- Diary quality indicators
- Instantaneous enjoyment related variables
These Stata Do files can also act as a guide to analyse other datasets with time-diary data.
Go to the time use and wellbeing repository
Using the publicly available version of the Civil Service People survey from 2011 to 2021, we analysed staff wellbeing and how it has changed through the pandemic.
We have used a Stata Do file to clean and extract the department-level ONS4 data for analysis.
Go to the Civil Service wellbeing repository
In May, we published our analysis of wellbeing across local authorities around England, based on the ONS’ Annual Population Survey.
To create our report, we used a Stata Do file code for data cleaning and analysis.
Go to the wellbeing around England repository
Why have we done this?
It’s good research practice to share your code – making it publicly available makes it more likely that any errors that slipped through our internal quality assurance processes will be picked up. Publishing the library is an example of our values in action: to be independent, practical, iterative and open.
It also lowers the costs of each new piece of research in this area, which we hope will lead to more research being conducted.
As well as a record of the past, and a resource in the present, this code is also a link to the future of research in this area. All the code is available for free, but we ask that if you make use of it, you cite the technical report accompanying it. Doing this means that we’ll be able to see who’s doing wellbeing research with this data, and we’ll be able to ensure that your work feeds into our future reviews and policy recommendations.