Psychological Safety: An Exercise

In early 2024, PlasmaPy maintainer Nick Murphy asked me to help facilitate a workshop for their project. This "summer school" would bring together community members of all kinds, from undergraduates to faculty, from experienced open source contributors to people who barely knew what open source was.

The diversity of attendees meant we had to be especially thoughtful about the kind of learning environment we wanted to nurture. When Nick suggested we do a whole session on psychological safety, I enthusiastically agreed.

The session ended up with four parts:

First, we introduced the concept of psychological safety, drawing on definitions and graphics from Timothy Cark, Royce James, the Coast Guard Spectrum program, and Kimberly Young-McLear.

Second, we talked about how safety is not defined by the absence of conflict, but the ability to handle conflict well, with warmth and belonging. We asked attendees to brainstorm a list of needs and expecations they were bringing to the workshop, as well as ways those needs and expectations might conflict. As facilitators, we also added our own needs and conflicts to the list—more on that below!

Third, we asked attendees to tell us about times they felt psychologically safe, and ask what about the situation or environment helped them feel safe. This ended up taking much longer than we expected—participants discussed their experiences for over half an hour.

Finally, we applied what we'd learned about psychological safety and managing conflicting needs to the process of pull request review. We asked what they, as new contributors, needed and expected from the pull request process, and how that might conflict with what maintainers need.

You can view the slides from our session here. But the core of the exercise is not the slides. It's the dynamic you create in the room.

During the second part of the session, I asked attendees to share needs and expectations they had. As most teachers know, these kinds of interactive exercises require audience participation to be successful. But people weren't raising their hands to share.

My instinct was to just move on. I was disappointed my activity was a dud, but there was no reason to dwell on it.

Except...my disappointment was itself an unfilled expectation.

I took a breath and said, in as lowkey a manner as I could, "There's actually a conflict happening in this room right now! I have a desire for this exercise to go well, with lots of participation. But you also have needs, like maybe you're feeling shy, or bored, or distracted. Those needs are in conflict with my needs. And that's okay! Conflict is okay. There's nothing wrong with my needs and there's nothing wrong with your needs, even if they are in conflict."

We moved on to the next part of the session, talking about times we'd felt psychologically safe and what had helped us feel that way. And maybe this would have happened regardless, but: people opened up. They got vulnerable with each other. For example, younger folks shared their apprehensions about working with mentors, and some of the faculty and post docs in attendance talked about their fears around failing their mentees. This part of the session ran for over half an hour.

It's hard to quantify psychological safety. It's hard to say how much one moment or one session contributed to the overall feel of the workshop.

But it's worth pointing out—what I did when I named the conflict in the room was take the exercise out of the realm of abstraction, from the informational level to the relational. I wasn't just telling attendees about conflict, I was demonstrating in real time that they could be in conflict with me and it would be okay. I wouldn't get angry and escalate, nor would I fear and avoid the conflict. By not ignoring the conflict, I let them know that they could trust me to raise potential issues rather than letting them fester and boil inside of me—another important element of psychological safety.

Days later, in the waning hours of the summer school, we were chatting with a few of the participants about their experiences and what could be improved. One person, an undergraduate, spontaneously said that they'd liked the psychological safety session, and in particular the moment when I'd named the conflict between my desire for participation and their hesitations.

Another attendee chimed in: "Yeah, when you said you had a conflict my heart started beating really fast, like did I do something wrong? But then you were like 'it's okay that there's a conflict' and I was, like, 'yeah, you're right, it is okay'."

I love that this person named their beating heart. Building a project together, being in community, learning from each other—these are all deeply felt things. They are experienced in the body. When we fail to acknowledge the relational, the embodied, it doesn't disappear. It submerges. It influences us in ways we have no way to understand or communicate to each other, because we've ignored and suppressed it.

Simply sharing our slides would risk missing this vital element. Some things can't reduced to handouts or procedures. Creating psychological safety requires demonstrating trustworthiness in real time, and that's not something you can get by following a routine. You have to show people that you truly care about their needs...and your own.