View/Export Results
Manage Existing Surveys
Create/Copy Multiple Surveys
Collaborate with Team Members

OR

Sign inSign in with Facebook

Sign inSign in with Google

Survey questions about diversity, and how to ask who is answering

Twelve questions about how work actually goes, plus five optional questions about who is answering, with categories copied from the published US and UK national standards. Answer it below, copy any block, or open the whole thing in the editor.

17questions
5 minto complete
5optional identity questions
2national standards named

Use this template Take the survey first

DEI Survey · answer to move through it

Use this template

Question 1 of 17 Respect day to day

I am treated with the same respect as everyone else on my team.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 2 of 17 Respect day to day

I can be myself at work here.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 3 of 17 Respect day to day

In meetings, what I say gets the same attention as what anyone else says.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 4 of 17 Getting on

Interesting work is shared out fairly here.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 5 of 17 Getting on

I can see people like me in senior roles here.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 6 of 17 Getting on

My manager backs my development as much as anyone else on the team.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 7 of 17 When something goes wrong

I know how to report unfair treatment here.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 8 of 17 When something goes wrong

I would report unfair treatment here without worrying what it would cost me.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 9 of 17 When something goes wrong

I trust this organisation to act when unfair treatment is reported.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 10 of 17 What is said and what is done

If I needed an adjustment to do my job, I could ask for one here.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 11 of 17 What is said and what is done

What this organisation says about fair treatment matches what I see day to day.

1

2

3

4

5

Strongly disagreeStrongly agree

Question 12 of 17 In your words

What is one thing that would make it easier to be yourself at work here?

Question 13 of 17 Optional: about you

What is your race and/or ethnicity? Select all that apply.

American Indian or Alaska Native

Asian

Black or African American

Hispanic or Latino

Middle Eastern or North African

Native Hawaiian or Pacific Islander

White

Prefer not to say

Question 14 of 17 Optional: about you

If you would rather describe your race or ethnicity in your own words, write it here.

Question 15 of 17 Optional: about you

Is the gender you identify with the same as your sex registered at birth?

Yes

No

Prefer not to say

Question 16 of 17 Optional: about you

Which of the options below best describes how you think of yourself?

Heterosexual or Straight

Gay or Lesbian

Bisexual

Self-describe

Prefer not to say

Question 17 of 17 Optional: about you

Do you consider yourself to have a disability or a long-term health condition?

Yes

No

Prefer not to say

Why this template has two blocks, and what the second one costs

Twelve questions about how work goes, and five about who is answering. The second block is the reason the first one is worth reading, and it is the one you localise before you send.

A diversity survey that asks nothing about the people answering gives you one company-wide number. That is a mood reading. It cannot tell you whether the experience here is the same for everyone, which is the question you were asked. Splitting the answers means asking, and that is what the second block does.

Asking has a cost, and it is arithmetic rather than policy. One identity question makes several small groups at once; two of them crossed makes a grid of them. What you can publish is decided by your smallest category rather than by your headcount, so do that sum before you send. If you would rather ask nothing about who is answering, an anonymous employee survey drops the block entirely.

Some of these answers are sensitive. Before the block, say in one line who will see the answers and what you will do with them, and let anyone skip any item. All five are optional and all five print an opt-out, which is what the published guidance says rather than a house style.[9][5]

A DEI survey in two blocks: twelve experience questions in four groups above a separate optional five-question identity block, and one whole-company result splitting into five groups of very unequal size — The identity block turns one average into a comparison. It also makes some of the groups too small to publish.

This page starts where the method page stops. Which topics to include, how to word the options no standard covers, and where the block belongs are covered in how to write demographic answer options. What follows is the part nobody publishes: the categories, and the standard each list came from.

Placement is the one thing the guidance will not settle. The State of Missouri toolkit says there is no right answer, then gives both cases: first when the survey is aimed at one specific group, and last when people are wary of the questions.[9] This template puts it last.

The diversity survey questions, and what a bad answer looks like

Copy one block or all seventeen. The line under each question says what to conclude when that answer comes back badly, and every rated item is written to be read split by group rather than as a company average.

Plain text, one question per line, with the answer scale.

Respect day to day (3 questions)

The ordinary texture of a week. As a company average these say almost nothing; split by group they are the point of the survey.

I am treated with the same respect as everyone else on my team.5-point agreeBad answer: a healthy average with one group a point below it. That gap is what the survey exists to find.
I can be myself at work here.5-point agreeBad answer: low from people who rate question one well. Being treated correctly and not having to perform are different things.
In meetings, what I say gets the same attention as what anyone else says.5-point agreeBad answer: a wide spread rather than a low average. That points at particular rooms, not at the organisation.

Getting on (3 questions)

Fair treatment as it shows up in who gets the interesting work and who gets promoted. A group difference here leaves a paper trail you can go and check.

Interesting work is shared out fairly here.5-point agreeBad answer: low in one group. Go and look at who was staffed on what for the last two quarters.
I can see people like me in senior roles here.5-point agreeBad answer: low from a group your headcount says is there. The one item you can check against the org chart.
My manager backs my development as much as anyone else on the team.5-point agreeBad answer: low in a single team. That is one manager to talk to, not a policy to write.

When something goes wrong (3 questions)

In order: whether people know the route, whether they would use it, whether using it works. A low score on the third makes the first two worthless.

I know how to report unfair treatment here.5-point agreeBad answer: low across everyone. A communication gap you can close this month.
I would report unfair treatment here without worrying what it would cost me.5-point agreeBad answer: high above and low here. People know the route and will not use it, which is worse than having none.
I trust this organisation to act when unfair treatment is reported.5-point agreeBad answer: low across everyone. Whatever happened last time is the answer you are getting, so find out what that was.

What is said and what is done (2 questions)

The gap between the stated position and the lived one. A leadership team asks about these first, so they sit last, where the earlier answers explain them.

If I needed an adjustment to do my job, I could ask for one here.5-point agreeBad answer: low anywhere. If asking is the barrier, the adjustments policy is not reaching the people it was written for.
What this organisation says about fair treatment matches what I see day to day.5-point agreeBad answer: low straight after a public statement. What is rated is the statement, not the workplace.

In your words (1 question)

One open box, before the identity block, so people still have a reason to type.

What is one thing that would make it easier to be yourself at work here?Open text, optionalBad answer: named colleagues. Theme them before anyone else reads them, and cut anything naming a project, a shift or a date.

The identity block, 5 optional questions

Every one can be skipped and every one prints its own opt-out.[9][5] Four are taken from a published national standard, named on the question; the fifth is ours. Change the categories for your own country before you send.

What is your race and/or ethnicity? Select all that apply.Select all that apply, optional US SPD 15, 2024 revisionOne combined question, seven minimum categories, as many selections as apply: the current US federal standard.[1][3] It binds federal agencies, not private employers.
If you would rather describe your race or ethnicity in your own words, write it here.Open text, optionalA write-in anybody can use, rather than an option labelled other, per the Arizona checklist: choose self-identify or write-in.[7]
Is the gender you identify with the same as your sex registered at birth?Yes / No / Prefer not to say, optional UK GSS harmonised standardThe standard collects gender identity, not sex, and is written for ages 16 and over.[5] It offers a write-in beside the second option; add one if your form can carry it.
Which of the options below best describes how you think of yourself?Single choice, optional UK GSS sexual identity standardThe self-completion version, the only mode where the standard prints prefer not to say as a choice.[6] The fourth option reads self-describe where the standard prints other.[7]
Do you consider yourself to have a disability or a long-term health condition?Yes / No / Prefer not to say, optional Written for this templateA self-identification item, not a measure of what somebody can do. The tested measure is the Washington Group short set, written for censuses and surveys.[10]

The block somebody will ask you to cut. The five identity questions are the ones a senior reader wants gone, because they are the ones that make a result attributable to a group. Cut them and you have twelve questions and one average, which is the survey you already had.

Where the identity categories come from, and why they change at the border

Four of the five identity questions are not our wording. They come from published national standards, named below, and two countries answer the same question differently enough that you have to choose.

The United States has one federal standard, revised in March 2024. It requires race and ethnicity to be asked as a single combined question, offers seven minimum categories, and lets people select as many as correspond to how they identify.[1][2] The seven are American Indian or Alaska Native, Asian, Black or African American, Hispanic or Latino, Middle Eastern or North African, Native Hawaiian or Pacific Islander, and White. Federal agencies have until March 2029 to comply, so a form you meet this year may still carry the older shape.[3]

The United Kingdom asks the same thing the other way round. The Government Statistical Service harmonised standard asks for ethnic group in one question, asks people to choose one option, groups the categories under five headings, and files Arab under Other ethnic group rather than giving it a heading of its own.[4] It also says each UK nation has its own requirements, so the country version is the one to use, and it recommends asking national identity first.[4]

	United States	United Kingdom
Who publishes it	Office of Management and Budget, with the Census Bureau[1][3]	GSS Harmonisation Team, Government Analysis Function[4]
Race and ethnicity	One combined question, required for self-response	Ethnic group in one question; national identity asked first
How many you may pick	Select all that apply	Choose one
Where Arab or MENA sits	Middle Eastern or North African, a category of its own	Arab, filed under Other ethnic group

The UK standards for gender identity and sexual identity print an opt-out themselves. Prefer not to say is in the government standard rather than something a survey writer thought to add.[5][6] One detail is worth copying exactly: the sexual identity standard prints it only in the self-completion version, because an interviewer records a refusal instead of reading it out as a choice.[6] An online workplace survey is self-completion, so that is the wording to take.

Even inside one country the published lists disagree. The federal standard writes Hispanic or Latino; two published university instruments both write Hispanic or Latinx.[7][8] One of the two carries Middle Eastern or North African and the other does not. One writes White or Caucasian where the others write White. The list is a decision, not a fact, and it is yours to make for the workforce you are asking.

One word to change before you send. Wherever you offer a write-in, label it self-describe rather than other, which is the Arizona checklist's rule against othering language.[7] That is why the sexual identity question here reads self-describe where the standard prints other. If your numbers have to match national statistics, put the published word back.

Two national standards side by side: seven categories with several selected at once on the left, five grouped headings with exactly one selection on the right — Same question, two standards: select all that apply across seven categories, or choose one from five grouped headings.

The UK version of the ethnicity question, to swap in before you send

What is your ethnic group?
Choose one option that best describes your ethnic group or background.

White
   English/Welsh/Scottish/Northern Irish/British
   Irish
   Gypsy or Irish Traveller
   Any other White background

Mixed/Multiple ethnic groups
   White and Black Caribbean
   White and Black African
   White and Asian
   Any other Mixed/Multiple ethnic background

Asian/Asian British
   Indian
   Pakistani
   Bangladeshi
   Chinese
   Any other Asian background

Black/African/Caribbean/Black British
   African
   Caribbean
   Any other Black/African/Caribbean background

Other ethnic group
   Arab
   Any other ethnic group

Prefer not to say (added here; the harmonised standard does not print an opt-out on this question)

Source: Ethnicity harmonised standard, GSS Harmonisation Team, Government Analysis Function. Use the version published for the UK nation you are asking in.

How to send it, and how to read the answers by group

Three decisions that belong to this survey. Writing the invitation, reminding and closing behave the same on any survey and are covered in the survey research methods guide.

Pick the cuts you can act on, and leave the rest out
Two identity questions you will use beat five you collected because they were on a list. Each extra one multiplies the small groups, and the toolkit puts the rule plainly: if the data is not needed, do not ask.[9]
Put one line above the block naming who opens the responses
One sentence in plain words, immediately before the identity questions rather than in a policy nobody opens. Name whoever reads the raw responses, and state what gets published.
Read the split before the average
The company number is the least useful thing this survey produces. Take the rated items cut by one identity question, then by the second, then read the open box against both.

Then let the answers pick what comes next. If what comes back is about how work gets done rather than who it gets done to, the organizational culture survey asks that directly. If the gap is about money, the pay fairness survey asks how pay is decided rather than how it feels. For the whole joining-to-leaving view rather than one cut of it, the employee experience survey questions cover the journey in modules.

All seventeen are already written and built, so you can open the whole thing as a free online questionnaire, swap the categories for your own country and send one link.

Diversity and inclusion survey FAQ

Do we have to let people skip the identity questions?

Treat it as settled. The State of Missouri toolkit puts it directly: self-identification should not be mandatory to complete the survey, and there should always be an option not to self-identify.[9] The UK gender identity standard says the same from the other side, that people should be able to skip the question and should be told at the start that they need not answer anything they do not want to.[5]

Which race and ethnicity categories should we use?

The ones published for the country you are asking in. The US federal standard asks race and ethnicity as one combined question, offers seven minimum categories including Middle Eastern or North African, and lets people select as many as apply.[1] The UK harmonised standard asks ethnic group as a single-answer question, groups its categories under five headings, and files Arab under Other ethnic group.[4] Copying one country's list into the other gives you numbers that line up with nothing.

Why is there no question about sex or about age?

Because neither has a settled list to copy, which is what this page is for. The UK gender identity standard says in its own words that it does not collect information on sex,[5] and sex question development and testing is still on the harmonisation team workplan.[11] Age is a different case: choose bands wide enough that your smallest one still holds people, and put the boundaries where a decision actually changes.

How small a group is too small to report on?

It is not the survey's number to set, and published thresholds disagree because they are written for different purposes. Fix your own rule before you send rather than after you have seen a result you would like to publish, and work it out from your smallest category rather than from your headcount.

Inclusion and environmental practice are both values questions and they are not the same question. For the environmental half, the sustainability survey questions cover practice, reporting and what staff can see happening.

Methods and sources

Survey methodology and questionnaire design, SuperSurvey

Survey methodologist and editor since 2003: wording, response scales, reliability checks, reporting that maps to a decision.

Every experience question written for this page
Every category list transcribed from the standard that publishes it
No Gallup Q12 item reproduced, reworded or matched in shape

Updated 7 September 2026 · How templates are reviewed

What this template is based on

The twelve experience questions were written for this page. The identity block was not: the combined race and ethnicity question comes from the US federal standard,[1][2][3] and the ethnic group, gender identity and sexual identity items from the UK harmonised standards.[4][5][6] The rules on optionality and on labelling a write-in come from a US state toolkit and a university checklist.[9][7]

Question licensing

Government statistical bodies publish these standards so that collections can be compared, which is what the GSS says they are for.[4] Nothing licensed appears, and no Gallup Q12 item is reproduced, reworded or copied in shape. The Washington Group short set carries no permission line, so this page names it rather than reprinting it.[10] Every question written here is free to copy and edit.

Limits and disclosure

A category list is a starting point. Published lists disagree inside one country as well as across borders,[7][8] each UK nation needs its own version,[4] and the US standard was revised in 2024 with agencies given until 2029 to comply,[3] so a list copied today needs checking before it is sent again. A self-identification question also measures what somebody is willing to tell an employer, which is not the same as who they are.

SuperSurvey makes the editor this template opens in. Everything here works with or without it.

References

US Office of Management and Budget. 2024 SPD 15 revision: revisions to OMB Statistical Policy Directive No. 15. spd15revision.gov
US Office of Management and Budget. SPD 15: question format. spd15revision.gov
US Census Bureau. Race and ethnicity standards updates, last updated 20 December 2024. census.gov
Government Analysis Function, GSS Harmonisation Team. Ethnicity harmonised standard. analysisfunction.civilservice.gov.uk
Government Analysis Function, GSS Harmonisation Team. Gender identity data harmonisation guidance, last updated 11 December 2024. analysisfunction.civilservice.gov.uk
Government Analysis Function, GSS Harmonisation Team. Sexual orientation harmonised standard, last updated 12 May 2021. analysisfunction.civilservice.gov.uk
University of Arizona. Inclusive and functional demographic questions, Assessment Toolbox, 2019. arizona.edu
Colorado State University, Division of Student Affairs. Demographic question wording suggestions. colostate.edu
State of Missouri, Office of Equal Opportunity. Diversity and inclusion survey toolkit, January 2022. oeo.mo.gov
Washington Group on Disability Statistics. WG short set on functioning, 19 March 2020. washingtongroup-disability.com
Government Analysis Function, GSS Harmonisation Team. Harmonisation team workplan. analysisfunction.civilservice.gov.uk

Send it this week, with the categories your country actually uses

Seventeen questions, about five minutes, and five of them anyone can skip. Open it in the editor, swap the identity block for your own country, and send one link.

Use this template