Who can use the data?
Currently academic researchers, government researchers, and non-governmental organisations may be able to use the LSs subject to approval. In order to use NHS data you will be required to complete additional application processes and training. Contact us for more information.
Will I have to pay to use the LSs?
The LSs are funded by the ESRC and Statistical Agencies in order to make them freely available to academic users. Non-academic research teams may be liable to be charged based on the time and complexity of the work being undertaken. If you are a non-academic researcher please contact us to find out if any charges will apply to your work.
How can I find out whether the LS has the data I need?
At what level of geography can I analyse the data?
The level of geography you can use for your analyses will depend on which LS you are using and on the variables you are interested in. Great care is taken to avoid small cell sizes which may allow identification of households or individuals. Even if you are allowed to use small levels of geography for your analysis there may be restrictions in the level of information you are able to see.
As this is a complex issue, if you wish to analyse data using low level geographies, we advise you to discuss this with your Support Officer.
Further information about geography in the ONS LS is available from CeLSIUS. SLS-DSU have published a Technical Working Paper (PDF 1.5MB) discussing geographies in the SLS. Information about geographies available in the NILS can be found on the NILS-RSU site.
For an informative presentation about census geographies and changes between census, see the UK Data Service Introduction to 2011 Census Geography
Can I link my own data to an LS dataset?
It is often possible to link your own data (eg, additional administrative or geographical data) to your LS project dataset. If you wish to do this, please discuss it with your Support Officer when developing your application.
Can I download the datasets?
It is not possible to download any LS data because of confidentiality restrictions and privacy agreements with the Statistical Agencies. However, techniques have now been developed to allow approved users to receive synthetic versions of their data for developing syntax prior to visiting a safe-setting. We also have downloadable versions of key variables in each LS which are useful for training and learning about using longitudinal data. For more information see our page about synthetic data.
Do I need to attend training before using the data?
Each LS requires slightly different forms and procedures, but all new researchers are required to have passed ONS Safe Researcher Training or Safe User of Research data Environments (SURE) training. Please refer to the following LS-specific links for more detailed information on what you will need:
- ONS LS – Using the ONS Longitudinal Study
- NILS – Becoming an accredited researcher
- SLS – Guide to accessing the SLS
How do you make sure the data are kept confidential?
All of the RSUs are strongly committed to maintaining the confidentiality of the individuals in our samples. Certain variables are not available for researchers to view, and security procedures in our safe-settings are strict. All outputs from analyses must be cleared by a member of RSU or Statistical Agency staff before being sent to researchers, in order to ensure that there are no small cell sizes or other potentially disclosive information included. More information can be found on the websites of each RSU or contact contact us
Which statistical software will be available in the safe-setting?
Most commonly used statistical software is available in the safe-settings, including Stata, SPSS, R and SAS. If there is a particular piece of specialist software that you require for your analyses, discuss this with your Support Officer and it may be possible to have this added to the computers in the safe settings.
Can I use more than one LS for my study?
It is not currently possible to combine data from more than one LS into a single analysis due to access arrangements which mean each dataset is only accessible within its own safe-setting. However, it is possible to either: conduct separate analyses on each dataset and then compare or combine the results; or use the eDatashield methodology which allows you to analyse data from more than one LS as though it is part of the same dataset. The eDatashield methodology currently enables all forms of General Linear Modelling, and we are working to extend it to other techniques such as Survival Analysis. Contact us to find out more.
How long does the application process usually take?
This will vary slightly depending upon which LS(s) you are interested in using.
- SLS-DSU advises that it may take up to 8 weeks to get initial approval after submitting your application, plus an additional 6 weeks for PAC application if health data are to be used.
- CeLSIUS advise that it should take no longer than 6-8 weeks for applications to be approved by the ONS.
- NILS researchers should not have to wait any longer than 2 months for their application to be approved by the Research Approvals Group at one of their regular meetings.