Measuring psychological safety

A quick, simple and interactive team exercise

What is psychological safety?

At Elsevier, one of our top priorities is to develop an inclusive culture built on trust, collaboration, & purpose. To achieve this, we need an environment where people feel psychologically safe.

How I measured psychological safety in my team

In order to understand if people in my team felt psychologically safe, I asked team members 7 simple questions: the 7 questions Amy Edmondson used in the study where she introduced the term “team psychological safety”.

  1. Members of this team are able to bring up problems and tough issues.
  2. People on this team sometimes reject others for being different.
  3. It is safe to take a risk in this team.
  4. It is difficult to ask other members of this team for help.
  5. No one on this team would deliberately act in a way that undermines my efforts.
  6. Working with members of this team, my unique skills and talents are valued and utilized.
A sample of how Mentimeter displays the results.

Questions to think about in advance

This exercise deals with a number of sensitive subjects — eg you might be asking people to share feeling excluded, rejected, or unvalued. So the exercise itself is best done in a psychologically safe space. If you’re thinking of running it with your team(s), I suggest you think about a few questions in advance:

  • How will you create the right conditions/environment in the room to discuss your results? The discussion could be very sensitive, so what guardrails will you have in place to set expectations and support people (eg a set of principles, rules of engagement, or a “safe space agreement”)?
  • How will you facilitate the discussion if your scores show that people feel a lack of psychological safety?
  • How will you facilitate the discussion if the majority of people score one way on a particular question and there is someone whose score is an outlier and very different from the rest?
  • How will you protect respondents’ anonymity?
  • How will you ensure that no-one is singled out?


I wanted everyone to answer the questions independently, without risk of bias/influence from seeing how other people had answered them. I mitigated the risk of this by not showing any of the results on the large screen I was using until I knew for certain that everyone had answered all the questions.

Things I’d do differently next time

The questions deliberately include a mix of positively and negatively worded statements to mitigate response bias. This is great from the point of view of ensuring the survey’s validity. However, it does make the survey quite complicated: 1 is not always a negative answer (indicating low psychological safety) and 5 is not always a positive answer (indicating a high degree of psychological safety). Next time, I would be clearer with people in advance that they need to read the questions particularly carefully.

Further reading/viewing


As I was digging around the internet, looking for other things to read on psychological safety, I came across this article by Martijn van Asseldonk, which says pretty much what I’ve just said above. I was surprised, but I’ve decided to publish this piece anyway, even though I realise that it’s not original. The more people who learn about this stuff, the better. And if this post ends up helping someone feel more psychologically safe, then that’s a great result. (FWIW Eva Offermans has also written an article on how to measure and grow psychological safety in teams, using a slight variant on Amy Edmondson’s questions.) A long time ago, I wrote my PhD on Andre Gide, and I’m reminded of something he wrote:

Senior Director @ElsevierConnect doing product strategy implementation & performance. Mainly writing about getting from A to B, & digital stuff. Personal acc’t.