COVID-19 harmonised guidance

This harmonised guidance is under development. These questions have been developed to collect data about the effect of the coronavirus (COVID-19) pandemic in a harmonised way. They have been adapted from government surveys. This process has involved changing the reference period and removing response options that were too specific. Some wording has also been changed. This is based on established questionnaire design principles and a consistent style guide.

We developed these questions quickly in response to the pandemic. This means that these questions have not been tested with the public in interviews. Instead, we developed the questions based on best practice principles and iterations with available evidence. This means that these questions are experimental and may change in response to new priorities.

We will continue to review this guidance to make sure it stays relevant and accurate as the situation develops. This work is set out in the Government Statistical Service (GSS) Harmonisation Workplan.

Policy details

Metadata item Details
Publication date:20 May 2020
Owner:GSS Harmonisation Team
Who this is for:Users and producers of statistics
Type:Harmonisation standards and guidance

What we mean by harmonisation

Harmonisation is the process of making statistics and data more comparable, consistent and coherent. Harmonised standards set out how to collect and report statistics to make sure they can be compared effectively across different data collections in the Government Statistical Service (GSS). By harmonising we can produce more useful statistics that give users a greater level of understanding about a topic.

We are proposing a harmonised set of questions to collect data about the effects of the COVID-19 pandemic. These questions have not been thoroughly tested, so they should be considered ‘under development’. The questions do not form a full harmonised standard.

What we mean by coronavirus

All coronaviruses cause disease in people and animals. The viruses can cause the common cold or more severe diseases, such as COVID-19.

COVID-19 refers to the ‘coronavirus disease 2019’. The disease can affect the lungs and airways. It is caused by a type of coronavirus. This set of harmonised questions relates to COVID-19 only, and refers to this as ‘the coronavirus’ or ‘COVID-19’ in line with the Office for National Statistics’ style guide.

Questions and response options: inputs

The harmonised questions on this topic are designed to collect basic information, for use in most interviewer administered and self-complete surveys.

The choice of variables is based on priorities identified across government and how appropriate they would be to harmonise. They are not designed to replace questions used in specialist surveys where more detailed analysis is needed.


Do you think you have ever had the coronavirus (COVID-19)?

Response options

  1. Yes
  2. No
  3. Don’t know


Have you ever tested positive for the coronavirus (COVID-19)?

Response options

  1. Yes
  2. No


Would you describe yourself as having ‘long COVID’, that is, you are still experiencing symptoms more than 4 weeks after you first had the coronavirus (COVID-19), that are not explained by something else?

Response options

  1. Yes
  2. No
  3. Don’t know


Have you ever received a vaccine(s) for the coronavirus (COVID-19)?

Response options

  1. Yes
  2. No
  3. Don’t know
  4. Prefer not to say


How worried, if at all, are you about the coronavirus (COVID-19) pandemic?

Response options

  1. Extremely worried
  2. Very worried
  3. Somewhat worried
  4. Not very worried
  5. Not at all worried
  6. Don’t know


Which areas of your life are being affected by the coronavirus (COVID-19) pandemic? Please select all that apply.

Response options

  1. My health
  2. My work
  3. My education
  4. My household finance
  5. My well-being
  6. My caring responsibilities
  7. My relationships
  8. My access to groceries medication or essentials
  9. Other (please specify)

Using these questions

Question placement

These questions can either be added to a wider selection of questions about the coronavirus, or asked on their own.

Types of data collection these questions are suitable for

These questions are based on variables used in both interviewer administered and self-complete survey modes.

Using these questions in the Welsh language

The questions were designed in the English language. At present we do not provide a Welsh language translation. This is because:

  • there is user demand for these questions across the UK
  • Welsh language testing has not been completed to ensure a translation is comparable and appropriate

Harmonised standards based on Census research have been tested in the Welsh language. This means that we can provide Welsh versions of them. If you are interested in using a Welsh language version of a harmonised standard that has not been translated, please contact the team at

Presenting and reporting the data: outputs

Work out the total number of unique “yes” values to output suspected cases of the coronavirus (COVID-19).

Work out the total number of unique “no” values to output suspected non-cases of the coronavirus (COVID-19).

Without testing, people who did not display any symptoms of COVID-19 are unlikely to know they have had the virus at all. These are known as ‘asymptomatic cases’. These people are likely to select “no” or “don’t know” in response to this question. In these cases, we suggest that you add unique “don’t know” values to output the number of people who don’t know whether they have had the coronavirus (COVID-19).

Work out the total number of unique “yes” values to output probable cases of the coronavirus (COVID-19).

Work out the total number of unique “no” values to output probable non-cases of the coronavirus (COVID-19).

Work out the total number of unique “yes” values to output likely presence of long COVID.

Work out the total number of unique “no” values to output likely absence of long COVID.

Work out the total number of unique “yes” values to output has received at least one vaccine.

Work out the total number of unique “no” values to output has received no vaccines.

Work out the total number of unique “don’t know” values to output those who don’t know if they have received a vaccine.

“Prefer not to say” is included in response options because of the potentially sensitive nature of this topic and the need to maintain privacy. This is consistent with good practice in data collection for sensitive topics. Work out the total number of unique “prefer not to say” to output those who would prefer not to say.

Work out the total number of each response option outputs levels of self-reported worry.

Work out the total number of each response option outputs self-reported levels of each domain impacted.

Only output responses under “other” once aggregated or coded to different domains. Do not publish free text respondents provide.

From use on government surveys we know respondents are including both positive and negative effects when responding to this question. Because of this, outputs from this question should be reported as areas affected, and not areas negatively affected.


Guidance for UK comparability

Wherever possible we aim to create questions that work for each of the four nations. But health is a devolved issue in the UK. This means that England, Northern Ireland, Scotland and Wales took responsibility for their own response to the pandemic and any policies relating to public health.

The approach to the way lockdown restrictions were eased across the UK varied, with devolved countries following their own strategies on measures such as testing, social distancing and extended households. This means that it’s important to think about whether survey respondents were aware of these differences between the strategies in each devolved country. You should also think about how knowledge of these differences may have affected responses and data. This is especially relevant when using the previously included questions on key worker status and social distancing. To view these questions, please email the team at

Different policies across the UK countries may also have affected comparability of the outputs from these questions. For example, the data may show a higher number of key workers because there was a broader definition of what a key worker was. Equally, the data may show higher levels of diagnosis because of different policies on testing.


Comparing survey and test data

Without testing, we cannot know exactly how many people have or had the coronavirus. There are many reasons for this. For example, survey questions which rely on self-reported symptoms will miss asymptomatic cases. This means survey data on the topic is an estimate, and variance is to be expected.

Survey questions are also unlikely to be comparable to test data except in studies that use both survey and biological data, such as the Coronavirus (COVID-19) Infection Survey.

One reason why survey questions are unlikely to be comparable to test data is that testing figures will miss cases. This is because polymerase chain reaction (PCR) tests were only provided to a subset of people because there was a shortage of them. PCR tests were free and widely available to people experiencing symptoms at one point during the pandemic. But now PCR tests are not provided free of charge. They are now only available to buy from shops or pharmacies.

Throughout the pandemic, self-testing lateral flow tests (LFTs) have been and continue to be available. Like PCR tests, LFTs were free at one point during the pandemic. But LFTs are now only available to buy from shops or pharmacies. The results of these tests are self-reported. This means some test results may not be reported and some cases may be missed.

There is also concern about the reliability of LFT results. It is thought that LFTs can produce false negative or false positive results, so data may not be accurate. Some workplaces expect their staff to do regular PCR tests or LFTs. This includes a large number of people who work in health and care occupations, which means testing data may be skewed by occupation.

Testing data now also includes antibody testing, which tests people for past COVID-19 infection. So data users must be mindful that current testing counts now include people with coronavirus antibodies.

Comparing prevalence data

When you are comparing data on the prevalence of COVID-19, it is important to understand whether the data you are looking at are reporting:

  • new cases
  • current cases
  • cumulative cases

Cumulative survey data relates to questions that ask whether someone has had the coronavirus at all, such as the current harmonised question on this topic. Cumulative data provides information that cannot be compared to questions asking about whether someone currently has the coronavirus.

The Department for Health and Social Care (DHSC) and UK Health Security Agency (UKHSA) have a live tracker for both cumulative and new cases. This data should not be compared to survey data that aims to achieve a representative sample. This is because the live tracker is based on testing data, which is only available on a specific subset of people.

Levels of prevalence will also vary based on levels of testing. You should think about this when you are comparing outputs and especially when levels of testing are known to vary between samples.

Further information

Removal of questions

Additional questions on key worker status and social distancing were previously included in this standard. But because of the constantly changing nature of the pandemic and current alert level they have been removed. This is because the relevance of these questions has decreased since they were first published. Questions about key workers and social distancing are more relevant in periods of high alert.

Alert levels are determined by how quickly COVID-19 is spreading. This is measured by the:

  • reproduction number (R) — a measure of the reproduction rate of spread
  • number of confirmed cases at any one time

Governments may decide to ease or tighten restrictions based on the alert level. This could include deciding to enter a lockdown.

You can ask for the previously published questions about diagnosis and symptoms, key worker status and social distancing by emailing the team at

GSS Harmonisation Workplan

The GSS Harmonisation Team published the GSS Harmonisation Workplan in February 2022. The team are based in the Office for National Statistics (ONS).

The workplan shows that we will continue to review the COVID-19 harmonisation guidance. This is to make sure that the information stays relevant and accurate as the situation develops. In line with the workplan, this guidance was reviewed in August 2022.

If you would like to be involved with this work, please contact the team at

Wellbeing and the coronavirus

Data from the Opinion and Lifestyle Survey found that the proportion of adults likely to be experiencing some form of depression during the coronavirus pandemic had almost doubled from before the pandemic. The data were collected between July 2019 and March 2020. 84.9% of the adults that were experiencing some form of depression reported feeling stressed or anxious.

If data collectors are interested in investigating this aspect of the pandemic, we would recommend using the harmonised standards on personal wellbeing.

Social isolation

An area of interest in relation to the coronavirus pandemic is social isolation. Social isolation is different to physical isolation. This is because it is possible for people to be physically isolated from other people, but still feel socially connected to them. The loneliness harmonised standard includes a question on frequency of social isolation.

For children and young people the question is:

  • How often do you feel alone?

The response options are:

  • Hardly ever or never
  • Some of the time
  • Often

Time series

Data can be compared over time to monitor change. This is sometimes called a time series. When there are changes in the way that data is collected we may say that the time series is ‘broken’. This means that the data before and after a point in time should not be compared to each other.

Because of the coronavirus pandemic, some surveys have had to change how they collect data. This includes moving data collection online, removing variables, or changing who is surveyed. All these changes mean that data collected during the pandemic may look different to data collected at other times. This means that the data should not be compared over time. Although the data should not be compared, this does not mean that one set of data is of a higher quality than another.

To reduce the effect that breaks in the time series have, data collectors can do a ‘parallel run’. This is where both the original and the new data collection methodologies are run at the same time in parallel. This provides data from both methods covering the same time period. These data can then be analysed for comparability, which can then inform decisions on how data can or cannot be compared over a time period. Without a parallel run, it would not be possible to know whether the data can be compared.

An example of managing a break in the time series comes from the ONS. The UK Labour Force Survey has been running in some form since the 1970s. But in the process of a wider data collection transformation programme the ONS have started research into how data can be collected online using a new, prototype Labour Market Survey. To understand the effect of this, the ONS completed a parallel run of core outputs and published a comparative estimates report. The report notes that headline estimates show no significant differences, but there were differences at lower levels. This might suggest that headline estimates can be compared over time but detailed analysis should not be attempted.

Relevant harmonised standards

We recommend using other harmonised standards that may be relevant in data collection during this time, such as:

You can help to improve harmonisation of statistics and data by using these standards in your work.

We also suggest looking at information that has already been published before you start to collect more data on the topic of the coronavirus. This includes:

Coronavirus (COVID-19) question bank

The ONS have created a bank of questions from surveys that ask questions related to the coronavirus.

Surveys involved in development

This question bank has been developed using the following surveys:

  • COVID High Risk Group Insights Study (CEV) ​
  • COVID Test & Trace (T&T) Cases Insights Study ​
  • COVID Test & Trace (T&T) Contacts Behavioural Insights Study
  • COVID Test & Trace (T&T) Self-isolation insights study
  • COVID-19 Infection Survey (CIS) ​
  • COVID-19 Schools Infection Survey (SIS) ​ ​
  • COVID-19 Vaccine Opinions Survey (VOS)
  • Household Assets Survey (HAS)
  • International Arrivals Insights Survey (IAIS)​
  • International Passenger Survey (IPS) ​
  • Labour Force Survey (LFS)​
  • Labour Market Survey (LMS)​
  • Living and Foods Survey (LCF)​
  • Opinions and Lifestyles Survey (OPN) ​
  • Over 80s Vaccine Insights Study ​
  • Student Coronavirus Insights Study (SCIS) ​
  • Student Experience Insights Survey (SEIS)
  • Survey of Living Conditions (SLC)​
  • Time Use Survey (TUS) ​

You can find more information about these surveys on the ‘About the studies’ tab of the question bank. The question bank will be updated regularly to include new questions and surveys.

If you have any questions about the question bank, please contact the team at .

Why we are sharing the question bank

We have two main intentions for sharing the question bank.

The first intention is to provide a list of questions to be used in other surveys.

The question bank and harmonised questions both cover similar topic areas, including the effects on life, health and social contact.

When developing a new questionnaire, we recommend that you use the harmonised questions first.

You should use the coronavirus  question bank if you need to harmonise a set of questions with a specific data source or assess how other surveys ask questions on topics not covered by the harmonisation guidance.

The second intended use of the question bank is to provide users with an understanding of what data the ONS has in relation to the coronavirus pandemic.

This will allow specific analysis from us to be requested as not all data from questions asked on ONS surveys are published.

You can find more information about how to ask for statistics or access to the research data on the ONS website.

You can find the latest publication from these surveys in the ‘About the studies’ tab of the question bank.


We are always interested in hearing from users so we can develop our work. If you use or produce statistics based on this topic, please contact the team at

Review frequency:

This guidance will be reviewed regularly.


Date Changes
18 October 2022 The guidance has been updated to reflect legislation and alert level. The diagnosis and symptoms, key worker status and social distancing questions, guidance and outputs have been removed due to lack of relevance since initial publication. In their place we have added questions about perceived COVID-19 past-infection, testing for COVID-19, long term effects of COVID-19, and vaccine status.
8 December 2021 The Excel document upload of the question bank has been removed and link updated to the new COVID-19 question bank dashboard.
24 June 2021 Added the Coronavirus (COVID-19) Question Bank, compiled by ONS.
23 October 2020 The symptom question on diagnosis has been updated in line with changes to the UK Government official symptoms list for the coronavirus (COVID-19).
2 October 2020 A question on social distancing has been developed, as has further guidance on wellbeing, social isolation, time series, and use in devolved administrations.
  • If you would like us to get in touch with you then please leave your contact details or email directly.
  • This field is for validation purposes and should be left unchanged.