For weeknotes this week, some field notes from completing a survey from the Office for National Statistics.
Field notes from completing the ONS Shape Tomorrow
My household had an invite this week to complete the ONS ‘Shape Tomorrow’ Survey which explains that it covers “a range of topics, including work, unemployment, retirement, education, health and caring”. Besides the census, I think it’s the first ONS Survey I’ve been invited to complete in a long time - so it was interesting to run through with an eye on the user-experience, and thinking in particular about debates over potential use of administrative data linkage in place of surveys, and about the work we did with Care Full last year thinking about the representation of care in official statistics.
User experience - finding and completing the survey
-
Searching for details of the ‘Shape Tomorrow’ survey throws up very little background: leading either to the ONS pages for completing the survey, or to Mumsnet and Reddit discussions of people unsure if it’s a scam, or worried about how government have their details. I had expected to find at least some more detailed information on how the survey data was used and which statistics if feeds into. I turned to Gemini to try and drill down and find additional sources, which suggested that Shape Tomorrow may include elements from a number of other more well-know surveys (Labour Force Survey) in order to have greater ability to link question answers (presumably linking questions on health, education, work etc.). I was really surprised though how little information there was on why filling in this survey was valuable - and that the main pitch was that I should complete it simply because it is important for ONS to be able to produce statistics (okay: in fairness this kind of works on me… but the forum posts I saw suggest that professional awareness of ONS and an intrinsic bias towards helping out a National Statistical Organisation (NSO) is far from universal.)
-
Presumably because ONS does not rely on consent as it’s legal basis for data collection, there wasn’t an initial consent stage explaining in detail how study information would be used, although there was some info saying data would be ‘de-identified as soon as possible’ and deleted after five years. The first questions asked for the full names of everyone in my household: without explaining why this identifying information was important (I’m still not sure if it was just to be used to identify these people to me later in the survey, or if that data is used in some way to verify or additionally link data?).
-
The survey structure needs responses from each person in the household: but this had to happen from the same letter/login. It felt a clumsy workflow when I could perhaps have been asked to provide contact e-mail for each household member and/or to get a link I could give to each to provide their own inputs. I was asked for an e-mail address at the end of the process: but why not ask for that up front and offer to send reminders about completing the survey if you end up needing to break-off from it to work on other tasks (quite likely as it gives no estimate of how long completion will take).
-
Some of the survey pre-amble stated that the information being collected can’t be discovered otherwise. Yet, many of the questions on employment, income and so-on felt like they should be possible to answer from existing HMRC and other government records (which this page suggested ONS do have access to). I realise there are legal complexities of data linkage - and issues with the independence of the NSO: but it feels like there should either be a clearer explanation of why it’s important to independently collect info on employment and income etc, or to develop linkage tools that can be put in the hands of survey respondents to bring in their data.
Where is care?
Last year I had the real privilege of working with Ruth and Hannah from Care Full as part of the Connected by Data Community Campaigns on Data cohort, supporting their exploration of a campaign to get care better recognised in national statistics. Completing the survey with this lens was striking. I was asked about my work over the last week, and whilst answering I only work part-time triggered the ability to indicate the reason as childcare responsibility this was the only time caring came up in the survey questions.
The absence was notable: when asked about hours (and minutes!) worked in the last week, there was no way to indicate the reason I had to take time out of work was because of caring responsibility as an adoptive parent - and dealing with the considerable stress of government delays renewing critical adoption support funding. Although as I completed the survey, a message popped up about the data supporting action to shape tomorrow in my local community - I was left feeling a very big part of the community had been missed.
I’m not sure if I’d indicated I was not working, or if I’d indicated I was caring for an adult, if I’d have received more questions about caring responsibilities: but as a parent of children with complex needs, I was left acutely aware that this particular statistical collection would do little to illuminate how caring responsibilities impact on both community and economy.
Feeding back
There was a box at the end of the survey to share feedback on the process. Just as I was thinking about writing something, the phone rang - and so I hit submit to have the survey complete. However, I’ve jotted this post down instead, and will share with the ONS.