The project team meets every week and the progress report looks normal. One week before delivery, the manager discovers that testing has failed, cross-functional requirements remain unresolved and someone questioned the schedule a month earlier.
“Why didn’t anyone tell me?” the manager asks. The team remembers a previous colleague being criticized as unprepared after raising a risk. Another person who questioned a commitment was told to find solutions rather than report problems.
This is a hypothetical scenario developed for this article. It illustrates a management difficulty: leaders want accurate information, but their earlier responses may influence whether people speak up next time.
The management perspective: speaking up involves interpersonal risk
Amy Edmondson’s 1999 research defined team psychological safety as a shared belief that the team is safe for interpersonal risk-taking. A study of 51 work teams in one manufacturing company found an association between psychological safety and learning behavior. The source is linked below.
That is not a guarantee that a pleasant atmosphere improves performance, nor evidence that every industry will show identical results. Here, the concept is applied to risk reporting and manager assessment: fear of humiliation, labeling or retaliation may filter information before it reaches a decision-maker.
Manage safety and work standards separately
Allowing someone to say “I don’t know” does not remove the need to prepare. Allowing mistakes to be reported does not mean every mistake is exempt from accountability.
Managers can separate two questions. First: “What is happening, and what information do we need to limit the consequences?” Second: “Why did this happen, and does it concern capability, resources, process or responsibility?”
If the first conversation immediately turns into a judgment about effort, people may protect themselves rather than fill information gaps. But reassurance without follow-up or accountability cannot establish reliable standards either.
A useful practical distinction is between an experiment that did not work, a preventable lapse in an established process and deliberate concealment or violation of a known requirement. This is a management suggestion, not a psychological test or a legal classification of responsibility. Treating all three simply as “mistakes” obscures important differences.
One response can change the next report
When an engineer reports unstable test results, the manager can first establish what has been verified, what remains a hypothesis and when a decision is required.
There is no need for a lengthy motivational speech. “Thank you for raising this now; let’s establish the impact” can keep the discussion focused on the work. It can still be followed by clear actions, owners and reporting times.
By contrast, “You couldn’t handle something this simple?” may teach others to delay reporting until they can solve the issue themselves. The information reaching the manager becomes neater, but later.
The point is not to suppress every expression of frustration. It is to separate emotion, fact-finding and accountability so that one reaction does not close an information channel.
Make risk reporting a process, not a test of courage
If only unusually brave employees will raise concerns, the reporting mechanism is not sufficiently reliable.
A project meeting can include a recurring question: “Which assumptions, if wrong, would change our delivery commitment?” Each owner describes what they know without having to arrive with a complete solution. The manager then decides what requires investigation and what risk is temporarily accepted.
Some matters need a private conversation. Employees should not have to challenge a senior leader in a large meeting to prove the culture is open. Complaints and personal-data concerns should follow appropriate procedures rather than being exposed to the entire team.
What happens after a report matters too. If a problem has been recorded ten times without a decision or response, people may stop raising it. Safety involves not only avoiding criticism, but knowing that someone will take ownership of the information.
Observe improvement without creating a performative metric
“No one disagreed this month” is not a useful measure of harmony. Requiring every employee to raise a fixed number of issues each week can turn participation into performance.
Instead, look for concrete changes: are significant risks reported earlier? Is there a response? Do the same causes recur? Are there substantial off-record concerns that never enter the formal process?
An initial increase in reports does not necessarily mean the team has deteriorated; previously hidden information may be surfacing. But it is not automatic proof of cultural improvement either. Examine the nature of the reports, the response and the eventual outcome.
If anonymous surveys are used, remember that people in small teams may still be identifiable. The information should improve how work is done, not be used to discover who rated a manager poorly.
Ask managers about behavior, not whether they value communication
Most candidates will say that communication matters. A more useful question is: “When did a direct report last challenge one of your decisions? How did you respond, and what changed afterward?”
Then ask: “What happened when someone raised a warning that did not materialize?” This helps distinguish a manager who accepts concerns only after they are proven right from one who can work with reasonable but uncertain risk information.
Standards also deserve attention. How does the person address repeated non-reporting, recurring procedural violations or deliberate concealment? A credible manager should explain both how they encourage speaking up and how they hold people accountable—not simply how approachable they are.
Practical uses for employers and candidates
When commissioning a management search with Talent Nexus, employers can include risk reporting and team learning in behavioral interviews instead of leaving “leadership” as an undefined requirement.
Candidates can also ask a prospective manager how the team handled its last substantive disagreement. A perfect story is unnecessary. What matters is whether the person can explain the process rather than merely describe the company as open.
No manager can know every problem in advance. A more realistic aim is to make problems visible while options remain and ensure that the person raising them knows what will happen next.
Research and scope
Edmondson, A. (1999). Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 44(2), 350–383. This article draws on the original study’s public abstract. The meeting questions, interview prompts and observation methods are practical suggestions, not the study’s measurement instrument.
Original research and full-text repository: Harvard DASH