Guidelines on using the ethics self-assessment process

30 March 2022
Last updated:
30 March 2022

How to use the self-assessment

How to use the self-assessment

We recommend that self-assessments are conducted as early as possible in the project timeline, as this will help to determine and ensure the most ethically sound route for research. We also advise that you revisit the self-assessment throughout the project lifecycle to ensure that any changes to the proposed project are considered in light of the ethical principles.

Although this framework is presented as a self-assessment, it need not be a process that you complete on your own. It is also important to remember that the self-assessment process is designed to consider the ethics of your particular project – therefore, it is still the analysts’ responsibility to ensure that the project satisfies all of the relevant legal requirements relating to their project. We therefore recommend that you discuss your research projects and/or self-assessment form with the following (where relevant and appropriate):

  1. Senior director/manager of your branch/business area/organisation responsible for the research project
  2. The relevant data owner(s)
  3. Any relevant legal and data protection experts within your organisation
  4. Where appropriate, any relevant Communications and Media relations teams/individuals

All completed self-assessments should be sent to the UK Statistics Authority’s Data Ethics team, at The Data Ethics team is available to review finalised self-assessments and support thinking through mitigations to minimise against identified ethical risks.

To help you navigate through the process we have included a user checklist at the end of this document.

Back to top

The self-assessment form

The self-assessment form consists of 3 main sections:

  1. Basic Information
  2. Weightings for sensitive research areas
  3. Item drop down selection and justification

Information and guidance for completing each of these sections is provided in the next three sections.

The self-assessment tool calculates the average ethical risk for the research project and provides you, as the researcher, with a suggested course of action based on a risk-based outcome. Whatever the course of action, you should share your completed self-assessment form with the Data Ethics team, at before proceeding with the project.

The suggested courses of action are as follows:

Low risk:

Project may proceed after confirmation from the Data Ethics team

Average risk:

Consult with the Data Ethics team to discuss actions to mitigate any highlighted risks before proceeding with the project

High risk:

Consult with the Data Ethics team. If risks cannot be mitigated then this project should be presented to NSDEC for a full independent ethical review before proceeding

Back to top

1. Basic information

This section consists of five parts:

  1. Project title;
  2. Project timeline;
  3. Project purpose;
  4. Research overview;
  5. Data sources.

Project title

Please provide the title for your project. Please make sure that the title is indicative of the project.

Project timeline

Please provide some details about your project timeline. This should include key dates such as the start and end date of your project, as well as any dates for dissemination activities (such as project reports and outputs).

Project Purpose

Please provide a short summary of the project’s purpose. This should include the following information (where relevant):

  1. Project partners and/or sponsors
  2. Research aims and/or research questions

Project overview

Please provide details of how the project will be completed. This should include the following information (where relevant):

  1. Methods proposed / how data are collected, used, processed, and shared
  2. The research environment where the project will be completed
  3. Plans for dissemination of research findings
  4. Any useful and relevant background information

Data sources

Please provide a list of data sources that this project utilises, along with what type of data this is (i.e. Survey, Admin, Social Media, Web Scraped etc). Please also provide a justification for each of the data sources that explains why this data is requires and how this supports the public good of this work.

For further information on ethical considerations when using different types of data, see our high-level ethics checklist for third party data, our guidance on location data and our guidance on the use of machine learning techniques.

Back to top

2. Weightings for sensitive research areas

There are four characteristics (listed below) which help determine the ethical risk of a project, with ‘ethical risk’ being defined as the perceived likelihood of negative consequences of unethical actions. To measure the differential complexities of various ethical decisions these have been included in the self-assessment as weighted measures.

Linking data can lead to useful insights and offers new opportunities for existing datasets. However, as information about a data subject is pulled together from different datasets, the risk of re-identification of the individual increases. Data linkage may be also perceived as profiling, and hence might not be publicly acceptable. This weighting also applies to projects that utilise already linked data.

Personal data means any information relating to a person who can be identified, directly or indirectly, from the information. This definition provides for a wide range of personal identifiers to constitute personal data, including name, identification number, location data or online identifiers. Sensitive personal data are special categories of personal data as defined in law. These special categories include personal data on racial or ethnic origin, political opinions, religious or philosophical beliefs, trade union membership, genetic data and biometric data (where processed to uniquely identify an individual), data concerning health, and data concerning a person’s sex life or sexual orientation. Due to the risk of disclosing the identity of data subjects, along with other personal information, it is important that researchers put in place additional safeguards. This is mandated by law (Data Protection Act 2018, and the UK General Data Protection Regulation).

There are particular sensitivities when using patient level clinical (health) data for research and statistics. Consideration needs to be given to the public acceptability of using such data and respecting patient confidentiality.

Could this research and/or its outcomes relate to individuals based on their protected characteristics. Protected characteristics are defined as age, disability, gender reassignment, pregnancy and maternity, race, religion or belief, sex and sexual orientation, as per the Equality Act 2010 (

Examples of groups that may be at greater risk of disadvantage are considered as groups of persons that experience a higher risk of poverty, social exclusion, discrimination and violence, including, but not limited to, ethnic minorities, migrants, people with disabilities and isolated elderly people and children, according to the European Institute for Gender Equality.

See our guidance on considering public views and engagement for research and statistics projects for further information on public acceptability.

Weights have been developed to account for these complexities in the self-assessment process and are applied to the overall self-assessment outcome. As legislation, regulation, and methodology around these areas evolve, these weights will be reviewed. Some weights may be adjusted, and new weight categories may be introduced.

On the self-assessment form:

If any of these characteristics are relevant to your project, please indicate this on the self-assessment form by placing a “1” in the corresponding cell on the form.

If you would like more information about how they impact the self-assessment outcome, then please contact the UKSA Data Ethics team, at

Back to top

3. Item scoring scales

The scoring scales

In this section, you are asked to assess your project against 22 items grouped against the six UKSA ethical principles. For all of the items, we ask you to respond to each based on a 3-point scale. To do so, each item has a drop-down selection where you are asked to select the most relevant option to your project. These options are affiliated with a score of 1,2 or 3. For all items, we also ask you to add a justification of your selected rating for each item. To assist you with providing this justification, the form provides prompts on the information that is expected from the justification, based on the selection from the drop-down options.

Where appropriate and justified, some items can be omitted when completing the self-assessment by selecting N/A, but again a justification is required as to why the item is not applicable to your research. The N/A function is only available for some of the items and is illustrated in this guidance as per the N/A diamond to the right of the 3-point scale below.


As mentioned above, to avoid responses that might indicate ethical issues being averaged out of the overall outcome, we have introduced tolerances against each item that is scored. You will therefore notice that when a statement is selected from the drop-down list, this will be highlighted red to indicate the tolerance limit. For example, the tolerance level for the public good item is set at the middle response, “Potential to achieve public good which requires further exploration”. This is because the public good should always be an integral part of the research aims and should be known prior to starting. Without these tolerances, a project could therefore achieve a “Low Risk” outcome, despite there being no clear public good. When these tolerance limits are reached or exceeded, researchers should consider appropriate actions to mitigate the ethical risk. If mitigations are not possible, researchers should also set out a justification as to why. These areas of the self-assessment will then inform a conversation between the research team and the UKSA Data Ethics team to understand whether there are any steps that can be taken to minimise identified risks, and/or whether this issue would benefit from independent scrutiny from the National Statistician’s Data Ethics Advisory Committee.

In this guidance, tolerance levels against each item are indicated by a black diamond around the corresponding level on the 3-point scale. In the example below, the tolerance limit is set at 3. This is the most common tolerance level.

The items

In this section, we provide guidance on how to consider your responses to each of the 22 items grouped against the six UKSA ethical principles. We also describe which items have the potential to be omitted where such a response can be clearly justified.

Ethics self-assessment framework

  1. Public good, Population coverage, Potential harm, Biases
  2. Direct identification, Indirect identification, Data security, Consent, Permitted use of data
  3. Validity, Standards, Training, Human oversight, New Technologies, Potential to realise benefits
  4. Established legal frameworks, Established legal gateways and agreements
  5. Public engagement, Public views
  6. Data curation and re-use, Sharing of methods and tools, Public access to outcomes

Back to top
Download PDF version (449.19 KB)