The ONS Longitudinal Study: how does it work?

 

by Nicola Shelton

Back in the late 1960s there was concern that policymakers had too little information about births and deaths: death certificates recorded only limited information and even the occupation of the deceased could be recorded inconsistently. Similarly it was impossible to use information from birth registrations to look at patterns of fertility – how were children spaced within families, for instance? And so the ONS Longitudinal Study was born.

The 1971 Census had recorded respondent’s date of birth – as opposed to age –  for the first time. And that allowed statisticians to record data on a one per cent sample of the population – all those born on four dates of birth which were, and remain, a closely-guarded secret.

The ONS Longitudinal Study now holds records for more than a million people, none of whom have any idea that they are a part of the study. It’s only possible to join through being born in the UK or through migrating into it, and it’s only possible to leave by dying or emigrating. The study also holds information for those living with its members, but it doesn’t follow them up in the same way from census to census.

Similarly, the information provided on birth certificates did not allow for studies of birth spacing. Although such data could be obtained from the General Household Survey (GHS), the total sample sizes were too small for detailed studies.

From the start of the study, census data was linked to official records on the childbirths, marriages and deaths among study members. Every year the LS database is updated with information which also includes causes of death, widowhood and cancer diagnoses.  

And as the years have progressed, a greater variety of information from the census itself has been added.

So from 1971 the participants’ census responses on employment, education, marital status, area, housing and household composition were included, and this remained the case for the 1981 Census. In 1991 responses on ethnicity and life-limiting illness were added, and in 2001 this was supplemented by respondents’ reports of their general health and religion, as well as whether they were a carer. In 2011 languages spoken, national identity and passports held also became part of the census and were added to the study. 

What can the LS be used for?

Academics, official statisticians and others can use the data to find answers to a wide variety of questions. For example: are there links between employment status and mortality? If we live in a poor neighbourhood, are we more likely to be diagnosed with cancer, or to die of ischaemic heart disease? Does living near a power line make us more likely to get cancer? How does our mode of travel to work relate to the length of our lives? 

The unique nature of the LS, which enables information to be gathered on a very large sample of the population over time, also facilitates types of research which couldn’t happen otherwise. For example, researchers using the data can look at patterns of how people migrate after retirement; they can examine the effects of divorce and remarriage on the type of housing people live in, and they can compare individuals’ life histories through their education, their employment and their house moves. They can examine types of social change – for instance, are women having longer gaps between babies, and are parents more or less likely to survive cancer?

It’s also possible to use the data to look at certain ethnic groups or those living in particular types of area. The LS covers England and Wales and it has sister studies in Scotland and Northern Ireland, so the whole of the UK is covered. And it’s possible to make international comparisons, too: other countries including France, Denmark, Finland and the USA have similar programmes.

How can I apply to use it?

Although access to the data is restricted and can only take place within safe settings or by sending syntax to support officers to run remotely, the process itself is straightforward: 

1. Formulate your research question.
2. Contact the team to discuss your proposal.
3. Work with support officers to complete an application.
4. Submit the application.
5. Attend training.
6. Work in a safe setting on your analyses.
7. Receive cleared output by encrypted email.
8. If you want to publish, seek clearance before publication.

Both researchers and projects need approval before data access is allowed, and final checks are needed to make sure that any resulting publications don’t inadvertently identify members through using small numbers. 

For the purpose of preliminary work and for teaching, it’s also possible to obtain ‘synthetic’ datasets which contain very similar information to the actual data but without risk of any security breach.

How is privacy ensured for the study’s members?

In April 2020 the ONS launched a new  Research Accreditation Service (RAS) – an online one-stop-shop for becoming an accredited researcher under the Digital Economy Act, and accessing research data held by the ONS Secure Research Service (SRS) and other processors accredited under the Act.

The SRS uses a set of principles known as the Five Safes Framework for secure labs such as those run by the LS, designed to provide complete assurance for data owners. The Five Safes are:

Safe people: Trained and accredited researchers are trusted to use data appropriately.

Safe projects: Data are only used for valuable, ethical research that delivers clear public benefits.

Safe settings: Access to data is only possible using secure technology systems.

Safe outputs: All research outputs are checked to ensure they cannot identify data subjects.

Safe data: Researchers can only use data that has been de-identified.

What are the limitations of the LS?

Although the LS can reach areas most longitudinal studies cannot, it does have its limitations. When compared to cohort and panel studies it has a limited set of questions asked, and in some cases the definitions and details have changed over time. 

The main weakness of the LS is the lack of behavioural data – because it’s based on the census, it doesn’t ask the detailed questions about lifestyle that other studies can ask. 

But the LS is a powerful tool for researchers: it’s by far the largest nationally representative longitudinal dataset in the UK, and it enables detailed analysis of our social geography which could not otherwise take place.