We ask frontline workers directly.

Who takes part

Three frontline workforces, and a separate survey of UK shoppers.

Workers find the survey through paid advertising, mainly on Meta, and choose to take part. They name their employer or trust so we can check their sector. This does not independently verify their employment.

What we ask

Their experience, in their own answers.

Understanding a finding

What does 82% mean?

Each percentage describes the people answering that question. Some questions apply only to particular groups, and some people stop before the end. The number answering therefore varies.

What the labels mean

The terms used on the findings pages.

Verbal abuse

Being verbally abused, insulted or shouted at by a customer, patient or relative.

Threatened or physically attacked

In retail, one question covers both. In healthcare they are asked separately; the combined figure counts anyone who said yes to either.

Required to report

The worker says their employer requires incidents of aggression to be reported.

Nothing beyond an acknowledgement

The incident is logged and nothing follows, the worker gets an acknowledgement and no more, or does not know what happens.

Very, somewhat, not very, not at all supported

The worker’s own rating of the support they get after abuse. Unsupported means not very or not at all.

Negative about work

Answering “Dread it” or “Not great” to how going to work feels right now.

Within each level of abuse

Workers grouped by how often they report being abused and threatened in the past month, so that support can be compared between workers facing similar levels.

Combined figures

Where answers are added together, such as the three on The Public’s View, the total is worked out from the number of people who gave each answer and rounded once. Each part is rounded on its own, so the parts shown can add up to a slightly different figure.

Reading the findings

What to keep in mind.

People choose to take part

These are the experiences of those surveyed, rather than estimates for every worker or shopper.

People describe their own experiences

The answers draw on recall and judgement, rather than employer or police incident records.

Questions differ between surveys

Retail combines threats and physical attacks; healthcare asks separately. These figures are not directly comparable. Subgroup results say who they describe.

A relationship does not establish a cause

Support and feelings about work can be related without proving that one causes the other.

This is the first reading

Further findings are needed before we can describe a trend.