In 2025, we consulted with over 65 organisations across the public sector on how to improve the People matter survey.
While everyone told us they valued the survey as a vital tool for improving the workplace, we also heard we needed to:
As a result of this feedback, we:
Overall, clearer questions should make the 2026 results more accurate because they are easier to understand.
Because some questions changed, you cannot compare the 2026 results for 19 questions with previous years (or 25 if you work in a health service).
We have removed historical results for these questions to avoid misleading comparisons.
We did this if a question was affected in the following 3 ways:
Your reports tell you where we have removed results from previous years for questions.
Here is more detail on how we assessed this:
We reviewed wording changes to assess if the revised question was likely to change how people understood or answered the question.
We looked at if what we did:
We did not treat minor plain language or readability edits alone as enough to remove results from previous years.
We also considered the advice we received when we tested the survey with frontline and non-frontline employees in 2025, particularly where people understood questions differently.
We also compared 2026 results with 2024 to look at how big the change was.
We compared 2026 with 2024 because the same types of organisations took part in both years. We did not compare 2026 results with 2025 as:
If a result changed a lot compared with previous years, we treated this as a big enough change for us to look into it more.
A change in results alone was not enough for us to remove historical results.
We also looked at a range of other factors before deciding to remove results from previous years:
We have removed results from previous years for 19 questions (or 25 if you work in a health service).
We believe these changes had the biggest impact on the 2026 results:
People told us that the old survey questions made it hard to recognise negative behaviours such as bullying, discrimination and violence and aggression.
In the old survey, respondents were asked yes/no questions about whether they had experienced these behaviours.
For each negative behaviour question in the new survey, respondents are:
We made this change because people may not describe an experience such as bullying or discrimination unless they first see examples of the behaviour.
We also made it clearer that these behaviours can come from clients as well as colleagues and removed ‘unsure’ as a response choice.
You still get shown a definition for each negative behaviour on the page before you answer.
For example, we based our definition and response options for bullying on WorkSafe information. The bullying definition makes it clear that any behaviour must be repeated and unreasonable.
These changes align with the approach we already had for sexual harassment and were strongly supported by public sector organisations involved in redesigning the survey.
People told us they did not know what senior leader meant in the survey questions. Many people thought it was ‘just their manager’.
To make this clearer and easier to understand, we added a definition to each question about senior leaders.
For example:
In 2026, we have seen large changes to results, including that the ‘neutral’ score has gone up a lot.
If more people choose neutral, the positive result can go down even if the negative result stays the same.
We put questions in the survey into groups by theme for high-level reporting (for example, wellbeing, organisational integrity or engagement).
We also report an average score for all the questions in that theme (for example, an engagement score or wellbeing score).
As a result of the changes to the survey, in some cases we’ve had to:
Your reports tell you where we have removed results from previous years for question groups.
Find out more about impacted questions by downloading:
See results from previous surveys at: