BT

Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ

Topics

Choose your language

InfoQ Homepage Presentations Creating an Atmosphere of Psychological Safety

Creating an Atmosphere of Psychological Safety

Bookmarks
36:01

Summary

Tim Berglund discusses creating an atmosphere of psychological safety for a team and the impact it can have.

Bio

Tim Berglund is a teacher, author, and technology leader with Confluent, where he serves as the Senior Director of Developer Advocacy. He can frequently be found speaking at conferences in the United States and all over the world. He is the co-presenter of various training videos on topics ranging from Git to Distributed Systems to Apache Kafka.

About the conference

QCon Plus is a virtual conference for senior software engineers and architects that covers the trends, best practices, and solutions leveraged by the world's most innovative software organizations.

Transcript

Berglund: It's important whether you're a leader or an individual contributor, a formal manager or an individual contributor, this is the thing that you can drive, and I think that that matters. This is a part of an environment that you almost certainly want. You want this to be true of the place that you work, really, regardless of what you do. Even if you're not a developer at all, this is important for any team communication. It's just that as developers, we spend an awful lot of our time talking to each other. It's something that a lot of people from the outside don't appreciate. The part of your day that is heads down coding is usually not as long as you'd like it to be. There's a lot of talking with other people, even if it's not pointless meetings with marketing that you get dragged into, or whatever. We work together and debate things together. Creating an atmosphere of safety is key to making that collaborative process work well.

Safety

Again, this is a thing that we want to be a part of our lives. It's a thing that makes teams perform better. If you are a manager, you have a real responsibility to make this happen. It's up to you to make sure that the environment that your team works in is a safe one. If you're not a manager, this is still a thing you can contribute to. There's a lot of things about team culture you don't set, but there's behavior that you can engage in, that help makes things safe or not.

Google's Project

Google in 2012 embarked on what could only be described as a sociological research project, they wanted to know why some teams thrived and others didn't thrive so much. They wanted to know why some teams were successful and other teams weren't. They surveyed like 180 teams from all over the company. They're like Google, they think they're good at data, and they tried slicing and dicing that among as many dimensions as they could. They looked at that education level, whether the teams socialized together, gender diversity, ethnic diversity, whether they had hobbies in common in the team, all this stuff, and it didn't work. They couldn't really find any of those variables that controlled things.

Google's Frustration

This is a time to step back and think. We all do some of our work in isolation. You're heads down coding, and sitting crunching on something, you're in the zone. That's an extremely enjoyable experience for everybody who's in this profession. Again, that's not our whole day. As developers, we spend a lot of time just debating solutions, debating what to name something, debating maybe an architectural decision, maybe debating a big choice of the structure of a system or the adoption of a framework. They are consequential decisions. When we're working together in groups, there is magic that happens when teams work together. They exhibit a property called collective intelligence. This is this emergent phenomenon of a group of people where the group can achieve more together than its smartest individual could. Let's just say there's a spectrum of capability on the team. There's decisions and code that the best player on the team can output. When you're firing on all cylinders, you've got this collective intelligence thing going on, the whole team together is better than that person. Or, teams can be parasitic on intelligence. You've totally seen this, where a dysfunctional team actually produces worse results than its least capable member would individually. If you think about it, that's quite a trick.

Google's Discovery

After Google tried to find all these correlations, they did find something but it was completely behavioral. It had nothing to do with the properties of the people on the team. You can imagine if you were embarking on this research project, there might be certain things that you wanted to see. Maybe you wanted to think, teams that just hang out together, they do better. I just want that to be true. Or teams that are more diverse, are going to do better. I want that to be true. These were all dead ends. What they did find were two things, two properties. Separated the effective teams from the ineffective one. One was airtime management, which is where every team member felt like they had the ability to be heard in group discussions. It wasn't just one voice drowning out the others. It was, everybody really believed and experienced that they could speak if they wanted to.

We saw social sensitivity, which is defined as team members being able to sense how each other felt through unspoken cues. That's a tough one, because some people are naturally good at that, some people aren't naturally good at that. Teams where you had good airtime management, good social sensitivity, those were the high performing teams. All of the other things that they wanted to see or that they tried to see, didn't work out. This was a rediscovery of what a researcher named Amy Edmondson, had first named in 1999, this concept is old enough to drink. She called it psychological safety. She had her own take on this, of course, on how to put this into practice. I recommend, if you're interested in this that you follow up and you check out Amy Edmondson's approach. She's got a book on it. It's a little bit different from what I'm talking about here. She is the originator of this so she's worth knowing.

We came up with this definition that psychological safety is a team climate characterized by interpersonal trust and mutual respect, in which people are comfortable being themselves. That comfort is key to unlocking ideas to debate. In order to be creative and propose a new, potentially risky idea, people have to believe that the self that they're exposing, because they're exposing a little bit of themselves, some idea that comes from inside them. That's a little bit of the inner person that they're letting out. Normally that person is protected. They need to believe that that part of themselves that they're exposing isn't going to be attacked. Nobody really exposes all of themselves. There's always this outward person that you let people see, and then there's the inner thoughts. That boundary is in there somewhere for different groups of people and different levels of relationship, intimacy, and all that stuff. Certainly, your coworkers don't totally know the real you. If you protect it all, there's not going to be a creativity, there's not going to be any new ideas. If you let some of that be exposed in an environment of safety, then you will have better ideas.

Your Brain Under Threat

Some TED talk level neuroscience. What happens to your brain when it's under threat? There are a few things that happen. One is short term thinking. You don't think about what's going to happen long term. Your vision gets very close to you, and what you're able to imagine and see and think about gets very close to you when you're under threat. If you think about that, from a survival perspective, you don't want to think about the future. You want to think about the animal that's trying to eat you or the person that's trying to put a spear in you. Those are all very short term things. There's also diminished creativity. Creativity is expensive. It's not a time to be creative. That goes away when you're under threat. Black and white thinking. You don't think about nuance when there's danger, you think about this or that. This is related to a psychological phenomenon called splitting, where maybe you think of yourself as the best in the world or the worst in the world, or other people you might think of as the best or the worst. This is an actual developmental stage that young people go through as like adolescence. That's the way you can view the world. Normally, as an adult, you develop out of that. Under threat, you fall back to that, this very black and white view of the world and of other people.

Finally, you revert to pattern or revert to training. You don't think of new things, you fall back to exactly what you know. These are great and important, and we're made this way to be more survivable creatures. These are great things when we're under threat. When we're trying to design systems or debate code or debate naming schemes, the real things that we argue about, probably nobody's going to die. These are maladaptive to that environment. We don't want to trigger our brains into danger mode, when we're trying to make decisions together.

Techniques to Help Your Team Feel Safe - Managing Airtime

Let's go over some techniques to help teams feel safe. Managing airtime. This is one of the two key things. How do you do it? Again, if you're the boss, this is to you. If you're the team lead, this really is for you. If you have a flatter structure, and you're just an influential person, again, this is for you. It might be that you've got a place in the team where this is hard for you to steer. Let's just take a look at what the points are. You got to make space for quiet people. I'm a professional talker. This is what I do for a living. In meetings, you don't have to make space for me. If I have something to say, I'm going to take the microphone, I'm going to say it. Not everybody is like me. There's no rule that says that I'm the right way, and quiet people, that's just the wrong way to be. That's not true. Some people are quiet, and we have to make room for them to talk.

We have to help talkative people manage themselves. I've certainly had dementia team members who are also people who tend to grab the mic and want to talk. Sometimes you might have to take one of these people aside, maybe it's yourself. You have to develop some self-awareness about this. You have to develop the tools to know when not to talk. It's easy for you to talk, you've got good ideas. You're good at articulating them. You think the world's better off if they hear from you. Sometimes they shouldn't. You have to learn when to govern yourself. If you're a team lead, you may do some governing of people. You may have to say, "Tim, stop talking now. We've heard enough from you, other people need to talk." What you really need to help Tim do is develop the skills to govern himself. That's the growth that you want to see in talkative people.

Also, last point, we want to let people use their airtime to criticize your ideas, speaking to you, again, as a leader or an influential person, it's important that they be able to use their time to criticize you. That needs to be a safe thing. They need to experience over time that criticizing ideas in the room, particularly the ideas of the leader is not something that results in harm to them. They don't have negative consequences.

Social Sensitivity

Next, let's talk about social sensitivity. Again, this may be a thing that's completely natural for you, and nobody needs to tell you how to read the room, and you're super good at it. I'm the person who's good at reading a room, even to a fault. There's some interesting dark sides to reading the room that you have to look out for. If this isn't natural for you, here are some tips. First of all, just watch facial expressions. If it's really hard for you to read facial expressions, and I'm saying watch facial expressions, and you say, "Yes, Tim, I know. Everybody says that. I don't know what they mean." Sometimes that is a thing. If that's you, you can learn. It stinks if you have to, because it's really hard. You kind of, as it were, have to do this in software. If you feel like other people just do this. This is a feature of the hardware, and they don't have to think about it. They know what phases mean. You're like, "I never know what phases mean." You can do it. You just need to partner with a person who's got that hardware support. Somebody that you trust, who can tell you, this is what this phase means. This is what this phase means. You can learn it.

On Zoom calls, use gallery mode. Make sure you can see everybody's face all at the same time, because you're going to miss things if you don't. If you're just looking at the speaker, the speaker might say something and somebody else might be hurt or disappointed or put off. If you're not able to see all the faces, you won't know that. Of course, if you're in the room together, just look around, but typically, we're doing a lot of online these days. There's another little cue you can see when you're looking at those faces, which is watching for people who start to talk and then censor themselves. You got to look for that and give that person space. Say, "Look like you were going to say something, what's up? Everybody else shut up. This person is talking." You really want to clear the deck and make it welcome for them to speak.

Again, I know I'm saying stuff that you might be a total natural at this, and it is second nature and nobody ever has to tell you a thing. Or it might seem like completely impossible. I know just from friends I've had over the years who are in that, 'I have no idea how to do this category,' you can learn. It's like me and singing. I try to ask people, how do you make your voice do a certain note? They say stuff like, you just do. That's like completely unhelpful to me because I don't want to do it. It's not an ability I have. Apparently, most people can just do that. Ok, that must be nice. People like me can be taught, it's just laborious. Or like dancing. I've taken some dance lessons. I'm not a natural. I'm like a block of wood, but I can learn. I just have to grind it out a step at a time. It's never going to be great but you can get there.

Mirroring and Labeling

A way to help draw out emotional responses. If you think an emotion is being expressed, but you aren't sure what it is. There's this really easy technique, and it sounds stupid, but it works. What you do is somebody says something and you're like, "Ok, I think that was a feeling. I just don't know what the feeling was." Then repeat back the last three words that they said, whatever those words were at the end of the sentence. You could pick three other words if you want. You could just pick the last three words they said, just repeat that back. They take that as a question. People just interpret that as, she heard me and she wants to know more about that thing. It just sounds like this really smart, subtle question. It's bizarre that it works, but it works. Then they'll say some more, and if you still don't know, then pick the last three words, again, of the last thing they said. Keep probing like that. Eventually, you'll start to get an idea. You could say something like, it sounds like you are anxious about that, or it sounds like you are really enthusiastic about that thing. You come up with a label. The first thing is called mirroring. Then you come up with a label for the emotion that you think you're seeing.

The amazing thing is you can be wrong. If you're a person who has trouble reading emotions, you just do this mirroring thing, you get an idea, maybe you're really bad at it, and it's the wrong emotion. Then they'll tell you, because you've been doing this listening, you're bouncing back ideas at them. Then they know, you've put an idea there, and they can either say, "Yes, that's right," or, "No, it's more like this." If you do this, I know it sounds contrived, but this makes you seem like the best listener they've ever known. That listening, this mirroring and labeling is going to make people feel safe in expressing ideas. You got to trust me on this, it works. If you don't have a lot of emotion words, you can get a thing like this, you could just Google, emotion wheel. You'll probably come up with this diagram. You have super simple words in the middle. Then you go out and you get slightly more detailed emotion words. On the outside they're like, that might be stuff you don't even need at work. This can help give you words for your labeling.

Handling Big Feelings

Sometimes there are big feelings. Something might go a little bit wrong in a situation, or in a debate and somebody gets upset, angry, whatever, and there's just something big. This is key because usually safety is established in all of the little interactions, not in the big things. You have to get all these little interactions, do those well. The big ones can leave a mark too. How you handle big emotions can help establish the context of safety. There are two steps when somebody is real upset about something or just having a big and probably negative emotion. What you don't do is say for example, calm down. Nobody has ever calmed down by being told to calm down. In fact, usually, being told to calm down makes people a lot less calm. What you do is reflect and diminish. The reflecting is if somebody comes to you and they're angry, like, "Can you believe this? We were told that we were going to get to build this on Kafka and we're not allowed to, and I'm mad about that." You reflect that a little bit. You don't tell them to calm down. You don't tell them they're bad to be angry. You're like, "Yes, I can't believe that." Just reflect that emotion back a little bit, but a little bit less of it. If they're up here angry, you're here angry, a little bit less. What that does is it says, it's safe for you to be feeling this thing that you're feeling, and I'm going to bring us down a little bit, to a little bit more control.

Normalizing Failure

Another important thing, normalizing failure. You see this teddy bear back here? This is a real background. It's not fake. This teddy bear is from a place called Build-A-Bear Workshop. That's a business that's 24 years old now, founded in 1997. My girls are all grown up. My boys are all grown up. Spent a lot of money at Build-A-Bear over the years. The founder, Maxine Clark, created something called the Red Pencil Award. This is from something she learned from a teacher when she was a little girl. The teacher, one year, had this award called the Red Pencil Award that was a recognition for the most mistakes made in a weekly writing assignment. That was cool, and all, the catch being you couldn't get credit for a mistake you'd made before. What this did was this helped us celebrate failure where lessons are learned. She instituted this in Build-A-Bear, where you could get an award for pointing out mistakes. Not just failure, obviously, we don't want to celebrate that because that's a race to the bottom, but failure where we figured out why we failed. We can learn how not to fail that way again. You want to normalize that thing and make it in a word, safe and even rewarded to do. Some applications: buggiest release of the year, worst blog post of the quarter of engineering, contributes to the blog. This kind of thing. Just pick something and award that thing, but the person who gets the award has to understand what the mistake was and explain how not to make it again.

Safe to Grow

Also, you want to make it safe for people to grow and not stay in the role that they're in, especially if you're a leader. You want people to believe that it's good for them to outgrow you. That you want to see that happen. You don't want to say, this is the thing that you're good at and we'll just keep you here because you're good at this and you're not good at those other things. You want a growth mindset, not a fixed mindset.

Psychological Safety Operationalized

Some points to summarize, cap this off, how do we operationalize this? You ask these questions. I actually recommend this, if you want to try to measure how safe is your team, whether you're in charge or not. If you're not, you could ask your boss. You could say, I saw this talk and I want to do this. You ask these questions, and just 1 to 5 rating. Where 1 is strongly disagree, 5 is strongly agree. If I make a mistake, it's held against me. Members of this team are able to bring up problems and tough issues. People on this team sometimes reject others for being different. It's safe to take a risk. It's difficult to ask for help, and so on. No one on this team would deliberately act in a way that undermines my efforts. My unique skills and talents are valued and utilized. I'm free to grow in new areas of interest. Simply ask people what they think.

The catch is here, if it is an environment without safety, people might not be honest. If you sense that you really have a toxic, unsafe environment, asking these questions in a public and visible way is not going to work, so if you can make this anonymous, that might help. You got to watch, if you really sense, and I think you'll know, having thought about the concept of safety a little bit, you'll know if you're in an unsafe environment. It can be hard to break out of one without help from the outside. That's a culture that's been established by a leader and that leader is probably going to need a lot of help and a lot of work to be able to get out of that mode, and that might not be a thing that you as an individual contributor can do.

Questions and Answers

Van Couvering: One thing I did see as a common theme was this question of what techniques you have to help get their quiet people to talk. You talked about some of them, but particular people talked about how to do this in remote scenarios, for example, managing airtime, social cues, how has that changed remotely? How do you continue to encourage quiet people to come out when you're in a remote scenario?

Berglund: Encouraging gallery mode in Zoom, or Teams, or whatever, where you're looking at everybody's face where everybody's face is there, is a good idea. Encouraging video to be on. I would not recommend requiring video to be on. Make it a normal thing for that to be the case, and a good thing. You got to be careful in the way that you do that. Like I've heard some people encourage video to be on with very inappropriate sorts of statements. You want to make sure that it's a thing where you're saying, we just like seeing each other's faces, period, because it makes it easier for us to collaborate. Yes, looking at faces is key there. Somebody has to keep tabs. This is ultimately up to whoever is leading, formally. Somebody has to keep tabs on who's talking and who's not. If somebody is just not talking, then what I want to do after a little while is clear the deck for that person, just say, "Everybody else, hold up a second. I haven't heard from Neve, I would like to hear what she has to say or whatever." Just be explicit. Again, there's risk there. If it's a person who really didn't want to talk, that might be intimidating for them. To some degree, it is their job. We're working together, so if somebody really has a hard time talking in a meeting, then you got to work with them privately on that.

Van Couvering: Annie Ruda said that they've used whiteboarding sites to help draw out quiet people, because they can just drop a sticky note rather than trying to share in a group. You can just make your comment by writing something and sticking a note on the whiteboard versus having to speak. At the beginning of the meeting, says James, give them a heads-up that you will be calling on them or going around the room, and then give them a concrete prompt rather than the open ended, what do you think? Actually, I'm thinking of my son who cannot bear open ended questions, does not like having video on, and loves to use chat. I can see those techniques really working for someone like him.

Any thoughts on how to start from scratch when building a completely new team?

Berglund: That's a great opportunity, especially if you're the one building it, you get to set the tone. If you could keep these principles in mind, and just put them into effect, you can model the appropriateness of disagreement. You're going to find there's going to be somebody on the team who is going to be more likely and more willing to disagree with you. Again, these things all fall on a spectrum. There might be shy people who would think I could never do that, but there's going to be somebody who doesn't mind so much. Who sees things differently, and they don't mind telling you. That person is a gift. It's your opportunity to help build this environment of safety, because you can then model when that person disagrees with you that you're not a threat. You don't defend yourself. That's a good thing that there's now a new idea to debate. Lose some arguments. Be willing to lose. People are going to see this. Don't download these slides and go to that new team and say, "Here are all the things I'm doing, this makes you safe." It doesn't work that way, they have to experience it. You can be intentional about even acting out some of these things that they then get to experience, and they will intuitively know, even if they've never read a word that Amy Edmondson has written, or anything about Google's Project Aristotle 10 years ago. People know and experience this and behave accordingly, so just model it.

Van Couvering: It's funny, you personally have to feel psychologically safe to do the mirroring and labeling so that you can make the room more psychologically safe for others.

Berglund: Yes. That's a thing that anybody can do. That is leading that anybody does, regardless of your technical chops, your juice in architectural discussions, your place in the org chart, none of it matters. If you've got that in you, then you can lead out of that strength of yours. Even if you're feeling like you're faking it, you can do that mirroring and labeling which requires you to at least act like you feel safe. If you've got that strength to bring, you're leading.

Van Couvering: I've seen teams struggle maintaining that sense of safety when they're under a deadline. On the opposite side is, how do you help hold teams accountable and still maintain psychological safety? When there are constraints like that, how do you maintain that space while still having to meet schedules and maintaining accountability?

Berglund: That's when it gets tough, because that's when you feel the danger mechanisms are kicking in. Again, that's the responsibility of the leader. This is one reason why to lead really does require a lot of self-awareness about these things, and a lot of strength on the inside. You now are the one who's afraid because your team is going to miss the deadline, and that impacts you in a bad way because it's your commitment. You have to make sure that you're aware of what that brain threat mode stuff is doing to you. You can actually stop that in software. You can be aware of it and you can say, "No, I'm going to decide on principle to act differently." As long as you get out ahead of it and see it in yourself. You have to learn to look for those things in yourself and decide that you will behave otherwise, which is a tall order, I know. You're not always going to do it. That's a fact. This stuff is very real, and it will get you sometimes, but the more self-awareness you have, you have a shot at it.

Van Couvering: Do you have any resources for learning how to read facial expressions? I ask this as a neurodiverse person who has trouble with body language.

Berglund: I don't. It's exactly your situation that I was talking to, because I know neurodiverse people, you're like, what? I understand the things I just am like, what about that other people find easy. I don't have resources, but that's a good thing that I should.

Van Couvering: My son's 15, and he is engaging with a social coach. These social coaches are there for children and teenagers as well as adults to help them succeed better at work and relationships in general by teaching them. We're just getting started but like how to start a conversation, how to know when a conversation is ending, how to read facial expressions, all those things that are maybe obvious for some of us. Someone does math in their head, for neurodiverse people, they're like, "We need you to show the work on how you read that facial expression and that body language."

Berglund: Yes, because there isn't anything obvious about it. Those of us who are neurotypical just need to know that and be willing to be a friend and a help to people we work with that that doesn't work. Like if you're going to sing with me, I would need people to, number one, help me. Number two, be patient with me when I usually don't sound good, because I'm never going to sound good.

 

See more presentations with transcripts

 

Recorded at:

Sep 11, 2022

BT