DEI Survey Questions: 45+ Examples by Dimension

The best DEI survey questions ask people how they experience belonging, fairness, inclusion, safety and respect at work, using a simple agreement scale, with demographic items kept optional and results reported only for groups large enough to protect identities. This guide gives you more than 45 diversity equity and inclusion survey questions grouped by dimension, plus open-text prompts and an optional demographic section.

Adaeze Okonkwo, head of people at a 70-person logistics company, learned the second half the hard way. Her first diversity survey asked for gender, ethnicity, disability and tenure on every response. Two weeks later she opened the report and found a group with two people in it. Everyone in that office could guess who had written the harshest comment. Participation dropped the next time she ran it.

The questions matter, but the privacy design around them matters just as much. This article covers both.

Key Takeaways

• Group diversity in the workplace survey questions into nine dimensions (belonging, fairness, inclusion in decisions, psychological safety, manager inclusion, respect, career growth, accessibility, overall) so every result maps to an owner and an action.
• Use one 5-point agreement scale for most items and a 5-point frequency scale for conduct items, so results are easy to compare across cycles.
• Make every demographic question optional with a 'prefer not to say' choice, and collect only what you will act on.
• DEI results are sliced by group, and small groups identify people. FeedbackPulse keeps results locked until 3 completed responses exist and masks every department, location and employee-level slice below 3.
• Ethnicity, health and similar answers can be special-category data under GDPR, so check with your counsel before you collect them.

#What a DEI survey measures

A DEI survey is an employee survey that measures whether people experience fairness, inclusion and belonging at work, and whether those experiences differ between groups. It is a diagnostic, not a scorecard. The goal is to find where people feel the workplace works for them and where it doesn't, so you can fix specific things.

Most workplace diversity survey questions fall into three families:

• Perception items ask how people feel ("I feel I belong at this company").
• Experience items ask how often something happened ("How often have you been interrupted in meetings?").
• Awareness items ask whether people know a policy or process exists.

Perception items are the easiest to answer and the easiest to act on in aggregate. Experience items are more specific and better evidence. A good survey uses both, and keeps awareness items to a minimum.

Diversity itself, meaning who works for you, is usually counted from HR records, not asked in a survey. What the survey adds is the lived experience that headcount can't show.

#Scales and rules to set before you write

Pick two scales and stick to them. Consistency makes results comparable and keeps completion time short.

Four rules keep the survey fair:

1. Make every question skippable. People should never feel forced to answer a question about their own identity or about a colleague.
2. Ask about behavior and processes, not about individuals. Never ask people to name a person.
3. Avoid double-barreled items. "I am respected and have opportunities to grow" asks two things. Split it.
4. Keep the tone neutral. Describe experiences, not politics. The aim is a shared picture of how work feels.

Our employee engagement survey questions guide covers general survey design if you want to pair this with a broader pulse.

#DEI survey questions by dimension

Use the lists below as a menu. For a first survey, choose two or three items per dimension. The letter in brackets shows the recommended scale.

#Belonging survey questions

Belonging is the feeling of being accepted as you are. These belonging survey questions work best on an agreement scale.

1. I feel like I belong at this company. (A)
2. I can be myself at work without worrying about how I'll be judged. (A)
3. People on my team value what I bring, not only what I produce. (A)
4. I feel connected to my colleagues. (A)
5. I would recommend this company as a place where people like me can succeed. (A)

#Fairness and equity of opportunity

These items test whether people believe the rules apply evenly.

6. Everyone at this company has a fair chance to get the assignments that build their skills. (A)
7. Decisions about pay are made fairly. (A)
8. I understand how pay and bonus decisions are made. (A)
9. Performance is judged on clear criteria that apply equally to everyone. (A)
10. When I have raised a concern about fairness, I believe it would be handled seriously. (A)

#Inclusion in decisions

Inclusion is whether people can take part and be heard.

11. My ideas are given fair consideration, whoever suggests them. (A)
12. I am invited to the meetings and conversations I need to do my job well. (A)
13. People with different backgrounds and perspectives influence decisions on my team. (A)
14. Leaders ask for input from people who are affected before big decisions are made. (A)
15. In meetings, everyone gets a chance to speak. (A)

#Psychological safety

Psychological safety, a term Amy Edmondson introduced in her 1999 paper on learning behavior in work teams, is the shared belief that you can speak up without being punished or embarrassed. Inclusion survey questions in this block tell you whether people will use their voice.

16. I can raise a problem or disagree with a colleague without fear of negative consequences. (A)
17. It is safe to admit a mistake on my team. (A)
18. If I asked a question that showed I didn't know something, I would not be judged for it. (A)
19. People on my team accept those who think differently. (A)
20. I would feel comfortable reporting inappropriate behavior if I saw it. (A)

#Manager inclusion

Managers shape the day-to-day experience more than any policy.

21. My manager treats everyone on the team fairly. (A)
22. My manager actively seeks different points of view. (A)
23. My manager gives me feedback that helps me grow. (A)
24. My manager responds respectfully when I share a concern. (A)
25. My manager makes sure workload and visible assignments are shared fairly. (A)

#Respect and conduct

Use the frequency scale here, because you are asking what happened, not what people believe.

26. In the past six months, how often have you been interrupted or talked over in meetings? (F)
27. In the past six months, how often have you heard comments that made you uncomfortable? (F)
28. In the past six months, how often have you felt your contributions were overlooked or credited to someone else? (F)
29. In the past six months, how often have you been excluded from informal conversations or events that matter for work? (F)
30. I know how to report disrespectful or discriminatory behavior. (Y)
31. I believe a report of disrespectful behavior would be acted on. (A)

Questions about experiences of harassment or discrimination can surface serious allegations. Decide before launch who reads open-text answers on this topic and what you will do if someone describes an incident. The survey governance checklist covers access and escalation.

#Career growth and promotion fairness

32. I can see a realistic path to grow my career here. (A)
33. Promotion decisions at this company are fair. (A)
34. I understand what it takes to be promoted in my role. (A)
35. I have equal access to mentors, sponsors or stretch projects. (A)
36. I have the same chance as my peers to get training and development. (A)

#Accessibility and flexibility

37. The tools, systems and workspaces I use are accessible to me. (A)
38. I can get adjustments I need to do my best work without a difficult process. (A)
39. My work schedule or location flexibility is handled fairly compared with my peers. (A)
40. Meetings and events are planned with different needs in mind, such as time zones, caring duties, religious observance or physical access. (A)
41. I would know where to go to ask for a workplace adjustment. (Y)

#Overall

42. Overall, this is an inclusive place to work. (A)
43. Overall, I am treated with respect by the people I work with. (A)
44. I believe leadership is committed to making this a fairer place to work. (A)
45. I have seen positive change in how we treat people since the last time we asked. (A)
46. How likely are you to recommend this company to a friend from a background different from yours? (0 to 10)

Item 46 is a variant of the eNPS question. Our eNPS guide explains how to score a 0 to 10 item.

#Open-text prompts worth adding

Open-text answers carry the context behind the numbers, and they are also the biggest identification risk. Keep them to three or fewer, and tell people not to include names.

• What is one thing we could change that would make you feel more included here?
• Tell us about a moment when you felt valued or that you belonged at work.
• Is there anything about how decisions are made that you think is unfair? Please don't name individuals.
• What would make it easier for you to do your best work?

Warn respondents that free-text comments can reveal who they are if they mention specific events, teams or dates. A short note above the box ("Avoid details that would identify you or others") is enough.

#Optional demographic questions

Demographic items let you see whether experiences differ by group. They also create the biggest privacy risk in the whole survey. Follow three rules.

1. Every demographic question is optional, and every one includes 'prefer not to say'.
2. Ask only what you will act on. If you won't analyze by religion, don't ask about it.
3. Explain why you ask and who will see results, in plain language at the top of the section.

Typical items, each with 'prefer not to say':

• Gender (offer an open 'describe yourself' option as well)
• Age band (for example under 30, 30 to 44, 45 to 59, 60 and over)
• Ethnic or cultural background (use the categories your local regulator or census uses, so the data matches how you report elsewhere)
• Disability or long-term health condition
• Caring responsibilities
• Tenure band
• Work location or department
• Level (individual contributor, manager, senior leader)

Each question you add multiplies the number of small groups. Gender by ethnicity by department by tenure can narrow to one person quickly, which is the next problem.

#Why small cohorts break DEI anonymity

DEI results are sliced by demographic group by design. That is the point: you want to know whether people in one group experience the workplace differently from another. But the groups you most want to hear from are often the smallest, and small groups can identify people.

Imagine a 40-person company with two employees who identify as having a disability. A chart showing their average fairness score is not anonymous. It is a report on two named individuals, and the other 38 people probably know who they are. Stack a second filter, such as department, and a cell of two becomes a cell of one.

Rafael Mendes, an operations director at a 120-person manufacturer, ran into exactly this. His survey tool let managers filter by site, shift and age band. A supervisor combined three filters, got a group of one, and saw a score of 1 out of 5 against the question about managers. Rafael pulled the filters the same week and rebuilt the report around larger combined groups. Nobody had typed a name, and the filters did the damage anyway.

#What FeedbackPulse enforces

FeedbackPulse's anonymous mode is built around a reporting threshold, not just the word anonymous:

• Anonymous survey results stay locked until 3 completed responses exist. There is no admin override, and an anonymous survey can't be converted to named after responses arrive.
• The same minimum applies to every slice. Each department, location and employee-level view is masked until that group has at least 3 completed responses, so a two-person team never resolves.

Three is a floor, not a recommendation for demographic cuts. The masking rule covers the slices the tool reports. It can't control what you compute or share elsewhere, so set your own higher minimum for any demographic breakdown you present. Our guide on survey anonymity thresholds compares thresholds of 3, 5 and 10 and explains why repeating the same filters every quarter can shrink a group over time.

#How to protect small groups

• Collect only what you'll act on. Each extra demographic question creates more small groups.
• Combine small groups. Report 'under 45' and '45 and over' rather than five age bands. Use 'people with a disability or long-term condition' as one group rather than splitting by type.
• Set a minimum group size in writing before you see results, and tell employees what it is. A rule you invent after seeing a small group looks like a decision about people.
• Report overall results first. Compare groups only when each has cleared your minimum.
• Watch repeated cuts. The same filter every quarter can narrow a group as people join and leave.
• Treat open-text with care. Remove names and identifying details before sharing comments, especially from small groups.

#Data protection and demographic answers

Some demographic answers are sensitive in law as well as in practice. Under UK and EU GDPR, data revealing racial or ethnic origin, religious beliefs, health, sexual orientation and similar characteristics falls into special categories that need an additional legal condition on top of your ordinary lawful basis. Rules on collecting ethnicity or health data at work also differ from country to country.

That makes it worth deciding up front whether to collect these items at all, and whether they should be collected anonymously. Our GDPR employee surveys guide walks through lawful basis, special-category data and DPIAs, and the survey governance checklist covers retention and access. This is general information, not legal advice, so check with your counsel about what applies where your employees are based.

#How to run the survey

1. Tell people why you are asking and what you will do with the answers. Name the sponsor and say when results will be shared.
2. Turn on anonymous mode and confirm the minimum group size before launch. Anonymous surveys in FeedbackPulse keep results locked until 3 completed responses exist.
3. Keep it short. Choose 15 to 25 questions. Leave out dimensions that you can't act on this quarter.
4. Give a window of one to two weeks and send one reminder. For a one-off survey, employees without a work email can take part through a shared link or QR code.
5. Check participation by group before you read the scores. Low participation in a small group is a signal that trust is low, not that everything is fine.

You can build your own question set with any of the lists above using the team survey builder.

#What to do after the survey

A DEI survey that goes quiet afterwards does more harm than not asking. People noticed that you asked. They will notice if nothing changes.

Tamar Levi, a people lead at a 55-person design agency, shared three results in the first week: the questions people rated highest, the two that scored lowest, and one commitment for each. The lowest-scoring item was about promotion criteria. Within six weeks she published the criteria for each level and ran a short pulse on item 34 to check whether people understood them. Scores moved, and so did participation in the next round.

A simple loop works:

1. Share the headline results with everyone within two or three weeks, including the uncomfortable ones.
2. Pick no more than three priorities. Each one needs an owner, an action and a date.
3. Separate perception from evidence. If people believe promotions are unfair, check the promotion data too, and share what you find.
4. Re-ask the relevant items in a short pulse after the changes land.

Our guide on acting on pulse survey results walks through turning scores into owners and deadlines.

#Conclusion

Strong DEI survey questions are specific, neutral and answerable without fear. Cover the nine dimensions, use two consistent scales, keep demographic items optional, and design the reporting so no group is small enough to point at one person. Then act on what you learn and say so.

If you want to try it, build the survey with the lists above and run it in anonymous mode.