Accessing SLS data

Because of the sensitive nature of the data held in the SLS, raw microdata files are not provided to users. Instead, two options for accessing SLS data are available: 'remote access' and 'safe setting'. Data access is described below.

Remote access

The most common form of data access (remote access) involves the creation of a subset of the SLS dataset in SPSS, SAS or STATA on behalf of the user, following extensive discussions about the design. A version of the database, which has had all the actual data removed except for variable names and labels will be released to the user. From this it is possible for users to specify analyses by creating syntax code which will be returned to the SLS team to be run on the original dataset. Assuming the code runs error-free, the results will be checked for confidentiality issues and, assuming there are none, will be returned to the user. Some restrictions on data will clearly need to be in place, as it should not be possible for individuals to be identified from the results of any analyses (hence, cross-tabulations, scatterplots and model residuals may be restricted in the output). This may become an iterative process as the user requests refinements to the analysis.

Safe setting

More rarely, for those who require more support, or who wish to use more specialist statistical approaches (such as multi-level modelling), there will be the option of working on a dataset in a 'safe setting' based in the General Register Office for Scotland (GROS) in Edinburgh. Such visits will be managed in accordance with strict rules protecting the data and only outputs from analyses will be released, once they are checked for confidentiality issues.

Step-by-step guide to accessing SLS data

In this section we explain how to access the data step-by-step. These steps are required for all projects, regardless of whether the data will be accessed remotely or in a safe setting. However, some pilot work to test whether a particular research question is feasible may be possible and this can be discussed with the SLS Support Team. Although a number of steps are involved for most projects, we aim to make this a smooth operation with the minimum necessary time delays.

Step 1: Formulate your research question

You should explore this website, and particularly the data dictionary, to see whether the research question you are interested in could be explored with the SLS. It might also be useful to investigate research that has been conducted using the England and Wales LS, as many of these studies could be replicated using the SLS; a publications list for the LS can be found at: http://www.celsius.lshtm.ac.uk/publications.html. Also, it is useful to check the list with SLS Projects to check whether the exact same topic has not been studied already (see http://www.lscs.ac.uk/sls/projects.htm). If you are not sure whether the SLS is suitable for your research, please discuss it further with a Support Officer.

Step 2: Contact the SLS Support Team

If you are interested in using the SLS please contact the team by emailing lscs@st-andrews.ac.uk, or by phoning 01334 463992. After discussing your project you will be allocated a Support Officer who will work with you to further develop your ideas. Your project will also be allocated a project number at this time.

Step 3: Attend a training session

We encourage all users to attend one of our training sessions, which are held twice a year. Depending on timing, it may be inconvenient for many users to wait for a training session before embarking on a project, so it is not obligatory to attend one of these sessions, prior to working with the SLS. For information on forthcoming events (both SLS training sessions and more general training in longitudinal methods) see: http://www.lscs.ac.uk/training.htm.

Step 4: Complete an application form

Once you have discussed the project in detail with your Support Officer you should download and complete an application to use the SLS. You must provide some details about you and your project, including a list of the variables you require. You may also wish to attach your own data to the SLS and this must be detailed on this form. The completed form should be emailed to your Support Officer. The form can be downloaded here: SLS Application form.
If you want to use hospital admission data, cancer registration data or maternity data, you also need to complete an application form for submission to the Privacy Advisory Committee (PAC). You can ask your SLS support person for an application form. Only if you have approval from the SLS Research Board and PAC can your project be carried forward.

Step 5: Project submitted to SLS Research Board

Once the final versions of the application and undertaking forms have been received by the SLS Support Officer the project will then be submitted to the SLS Research Board. This board assesses whether projects should be supported and it may provide some advice on how projects could be improved. The SLS Research Board includes representatives from the SLS Support Team, and the major suppliers of data, including the General Register Office for Scotland, NHS Scotland and the Scottish Executive.

Step 6: Complete a SLS undertaking form

The SLS undertaking form must be signed by you and any collaborators that you are working with. The form deals with disclosure, confidentiality and ethical issues relating to your project. This form must be posted to the LSCS and no project can commence without this undertaking. The form can be downloaded here: SLS Undertaking Form.
In addition, users should at this stage read the disclosure control protocol.

Step 7: Dataset creation

When the project has been approved, the Support Officer will begin the process of creating the user-specific sub-set of the SLS data. The creation of the dataset may require some discussions between the user and the Support Officer before the final dataset is finalised. It is also possible for users to provide their own data to be linked to the SLS dataset. Once the dataset has been created the Support Officer will inform the user.

Step 8: Analysis

As described above, the analysis of the data may involve 'remote access' or a 'safe setting' approach. Please read the Working Paper which describes this process in more detail: SLS Working Paper 1.

Remote access: Under the remote access option raw microdata are not supplied to the user. Instead, the Support Officer will create a file which includes variables names and labels and category names and labels .This file will be released to the user who will then be able to create a syntax file containing the analyses they require and, once returned to the Support Officer, this file will be run on their SLS dataset. Assuming the syntax runs successfully, the results will be checked for disclosure and returned to the user. To maintain the security of the SLS all outputs are encrypted before being sent out to users and the approved encryption software is Sophos Free Encryption. Users can download this software for free from the following website: http://www.sophos.com/products/free-tools/sophos-free-encryption.html. After exploring the results, the user may then wish to adjust the syntax and repeat this process.

Safe setting: The safe setting option will involve the creation of a user-specified file which will be held in Ladywell House (GROS) in Corstorphine, Edinburgh, where the SLS database is maintained. Users will be able to visit the safe setting and analyse data alongside a Support Officer. Results from these analyses may be released from the safe setting, once approved by the SLS Data Custodian. They will be encrypted and emailed to the user.

Step 9: Publishing / presenting results

When you want to disseminate your SLS results beyond your project team and named associates you must obtain final outputs clearance from the SLS Data Custodian and sometimes the SLS Research Board (SLS Clearance Form). Typical final outputs will be working papers, reports or journal articles intended for publication, presentations or abstracts. The SLS Data Custodian must clear all types of output and you should allow 15 working days for final outputs clearance although many outputs are cleared more quickly than this. The process for clearing final outputs reduces the risk of disclosure, ensures that the study and data are properly described and that the data have been used appropriately. Key criteria that will be considered are:

• The results displayed and the discussion concerning them do not raise confidentiality or disclosure issues;
• The SLS is described correctly;
• 'Source: Scottish Longitudinal Study' is added to tables and figures, where appropriate;
• You acknowledge the support of the LSCS using this disclaimer:

"The help provided by staff of the Longitudinal Studies Centre - Scotland (LSCS) is acknowledged. The LSCS is supported by the ESRC/JISC, the Scottish Funding Council, the Chief Scientist's Office and the Scottish Executive. The authors alone are responsible for the interpretation of the data. Census output is Crown copyright and is reproduced with the permission of the Controller of HMSO and the Queen's Printer for Scotland."

Once your final output has been cleared for public release you may disseminate it a number of times without resubmitting it for clearance, as long as tables or figures are unchanged and there is no new discussion of SLS processes.

Step 10: Notify the SLS team of publications

Please notify the SLS Support Team when a publication which draws on the SLS is published. We maintain a library of all published research that uses the SLS and it is vital for us to keep this up-to-date.

Website comments to lscs@st-andrews.ac.uk

Valid XHTML 1.0