That's it Google? Psychological Safety ?

That's it Google? Psychological Safety ?

The New York Times described Google’s finding about what makes teams effective as “surprising” . The article tells the story of how members of Project Aristotle analyzed huge amount of data, looking for patterns that explain why certain teams are more effective than others. Their first culprit, team composition, turned out to not explain anything. The researchers then focused on finding the group norm that predicts effectiveness. Unfortunately the data showed that the same group norm that made one team successfully, created an only average team in another instance. The common factor in successful teams was a high level of psychological safety, a team construct that describes how safe the team is to make mistake and talk about them. Once this was ‘discovered’, the next step is to develop guidelines in how to make teams psychological safe. The article continues to tell the story of one manager who is compelled by this topic, and is willing to try out something to make his team work better. He begins a meeting with sharing a personal story, something that scares him. This highlights his vulnerability, his weakness. By doing the first step he showed that he is not without fault, helping the team to comeĀ  together, and work better.

After I finished reading the article, I thought “so what?”. The concept of psychological safety has been published in 1999 by Amy Edmondson. I know scholars who love the concept. Don’t get me wrong, I do thing it is crucial in teams. But the results are not surprising. The researchers at Google were able to replicate the results and this is necessary in science to generalize findings. But the problem persists: How to create psychological safe teams? The story about the manager underlines the problem even further. This manager wanted his team to be better. He noticed that something was wrong and wanted to improve. But what if your manager wants your team to be effective, but doesn’t want to spent extra resources in creating a psychological safe environment?

Screenshot 2016-02-26 21.53.05

Let’s make it more concrete and imagine three scenarios:

    1. Your team is the best team ever. You laugh together, you cry together. You all got each others back. Until the manager enters the office. Suddenly everything changes. S/he leaves, and psychological safety is restored.
    2. What if you don’t want to talk about your private life with you team, but want to create a psychological safe environment?
    3. What if there is a truck burning on the high-way, it is 5pm rush hour in Los Angeles, and +50 people have to be rescued from a pile of cars before they catch fire? No time to chat about difficult moments.

We know that psychological safety is important. Some may see it as a touchy-feely, squishy topic, and rather stay away from it. It is hard(er) to measure than other team aspects, such as team composition. But we know that we need it, so let’s thing about way to create it, preferably on the spot.

Possible solutions:

  1. Find the rotten apple : Use the instrument Amy Edmondson created and turn it into a network question. The goal is to ask each team member how safe they feel talking about failures and taking risks in the presence of each other team member. The result is a network with ties representing how safe you feel when this person is present. Now you can look for clusters, for those that make you feel unsafe. Depending on the phrasing of your question, these are people with no incoming ties (if your question is phrased positive) or with many incoming ties (if your question is phrased negative). You got your rotten apple and can focus your effort on this person. For example, a peer or superior can model the behavior, in the hope that the ‘rotten apple’ will pick it up. Or, if lower levels of psychological safety are related to certain physiological signs, the ‘rotten apple’ could get a notification every time s/he makes people feel unsafe. This would be something for Syndio or Humanyze.
  2. Do the flight simulate test: Pretty easy and fun too. Just put your team in a flight simulator with pre-briefing and de-briefing to change their behavior. Do this once a week, increasing the complexity of the flight missions. Test psychological safety every 2 to 4 weeks and continue until the team is safe.
  3. In the case of ad hoc emergence care staff the best option is to have simulations when there is no emergency. A de-briefing would also help. Important here is to hire the right people. Don’t go necessarily for the technical expert, but for those that have good interpersonal skills.

If you are seeking something more concrete, apply the advice that follows from research at MIT:

  • Make sure everyone has an equal amount of speaking time (equality in distribution of conversation time)
  • Test team members on their social sensitivity, their ability to gauge somebody’s emotion based on their tone of voice, their expression, and non-verbal cues. Only promote those to managers who score higher than average on this test.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Close Menu