Skip to content

Validity of Safety Culture Assessments

Many are questioning the validity of safety culture assessments. One of the reasons, as Sidney Dekker commented, is the ambiguous meaning of “safety culture”: 

“No one term or mechanism can adequately explain the social order in which we live. We use safety culture to encompass any or all aspects of social life. Then we link them to organizational safety. Is it unrealistic to expect the concept to do any verifiable and meaningful explanatory work?”

The other factor as I will discuss later is the negative effect of anonymous feedback on organizational learning.

In an effort to forge new tools for organizational social and psychological assessment. My previous article proposed that social systems thinking could lead to higher success rates in change efforts. Culture is very real. It is one of the forces shaping social systems. However, I have found that it is more actionable to address other elements such as power, relationships, emotions and feelings. 

While it is true that culture sets the rules for social interaction, it doesn’t set the rules for emotions and feelings. Our biological nature does. Power relations, the importance of status and the fear of not belonging are the most frequent triggers for the emotions that cause relationship breakdowns. One example is when what you are saying triggers my emotions. I may look like I’m listening, but my emotions are blocking my understanding. This is often a problem in meetings. It doesn’t help to establish a norm to “Listen for understanding, not agreement,” if it’s not happening outside the meeting.

Sometimes I wonder why some folks leading large-scale change like mergers, acquisitions or restructuring say that aligning leadership and culture are the most important elements of success. Then they spend 98% of their time planning everything but how they are going to deal with the swirl of emotions and feelings usually present. Have you ever tried to bring the subject into the picture? In my experience most times everyone nods politely and moves on. Why? Because we fear there is nothing we can do about it.

Improving safety performance, innovation or quality isn’t as traumatic as a merger, but the same emotions and feelings are in play because we are human. In this article we’ll look at anonymous surveys and the validity of safety culture assessments, which are two of the most popular. Do they help or detract from change efforts?

I have been an avid user and promoter of safety culture assessments and team building surveys. I developed one of the first safety culture assessment as part of my masters thesis in 1993. You can read my research paper published in the Professional Safety Journal. I have grown disenchanted with the process due to the high psychological and financial cost. The typical corrective actions that emerge don’t change the underlying problems. 

Most organizations use surveys as the beginning of the change process. The collective involvement in data analysis, prioritizing, creating an action plan and follow through makes change possible. I have seen it work. My dissatisfaction is with many of the unintended consequences of the process. It doesn’t change the social interactions that are the cause of disengagement. I will describe how this happens in more detail later.

Why are anonymous surveys so popular?

Companies need measurable data to measure progress. They use anonymous perception surveys to capture the emotional aspects of organizational life. They also want to provide a venue for employees to give honest feedback without fear of retaliation. 

On the management side it is felt that receiving feedback privately helps managers feel less threatened and more open to accepting the data. In my experience, managers who get negative feedback, even privately, are emotionally triggered. Since they continue to hold their position, not knowing how to manage that anxiety can worsen relationships. One of the employees’ fears is that the boss will know who said what. This does not happen with managers who have higher scores. I think because their higher scores reflect that they solicit feedback and use it as a way of doing business.

In my early days I would share the results by department with the whole management team because I didn’t realize that I was shaming the ones with low scores. In my mind the managers with low scores could learn from those with high scores. 

Why wouldn’t someone with a low score want to learn how to improve his or her team’s engagement? It was one of my biggest lessons that learning cannot take place when people feel blamed or compared unfavorably to others.

So, while the anonymity provides some layer of psychological safety for employees and managers, it doesn’t build trust or open communication. If mistrust exists it was created by the way people are treating each other and communicating. The only way to regain trust is to change the quality of these interactions to create collaborative relationships. Can anonymous assessment processes create the interactions to improve relationships?

Why anonymous assessment process don’t tend to improve relationships

Anonymous surveys and feedback worsen a We/They dynamic. When the survey results come out there is very little discourse. Managers feel blamed as the ones responsible for psychological safety. Blaming them isn’t going to make things better. In reality the manager’s behavior is often the result of group dynamics or organizational pressures. To make it worse anonymity makes it possible to give negative feedback without taking responsibility for one’s contribution to the problem. This doesn’t further learning how to disagree constructively or how to handle the inevitable miscommunications that arise in daily life.

The problems, which gave rise to the negative feedback, can only be resolved as a collective effort because everyone will have to change. This is true even if those problems stem from an individual’s behavior. The individual may try to change but cannot if the social system will not allow it. 

The typical assessment process can identify that there is low trust. Unfortunately, it doesn’t identify the emotions and actions triggering mistrust at the individual level. Without dialogue, managers can apply their own definitions of what survey results mean based on their experience. Sometimes employees engage in the problem-solving phase of the assessment process. Empowering employees to implement their ideas raises engagement and interest in safety.

I have seen this process work many times, always with a relationship centered leader. The sad part is that when the leader who championed the process leaves, it ends. My upcoming book The Relationship Factor in Safety Leadershipcontains many of their stories. To address this problem my colleagues argue that’s why we need a safety culture. It is one way to anchor the safety values so that they are not personality dependent.

Culture does not direct emotions it only governs how they can be expressed.

I’ve come to understand why the concept of safety culture isn’t working for me anymore. Culture does not direct emotions it only governs their expression. This diminishes the value of culture assessments. They cannot help us understand emotion and social processes such as trust. Even uncovering some of the emotions through focus groups or interviews we can miss the mark with our corrective actions. This is mostly because they tend to be one time fixes, new rules or processes that increase bureaucracy. 

For behaviors to change, we must first change the way people interact. Otherwise the current conditions get reinforced with every exchange. For example, many executives say that a strong safety culture is one where employees will speak up to stop and unsafe actions. The healthcare industry spends billions on preventable medical errors that someone questioned, but did not speak up. I agree that changing these dynamics begins with the power holder. That is not enough in itself. The other players have to agree to question a superior, which feels like big risk. 

This approach doesn’t address all the systems that impact safety performance

It is not my intention to discount the organizational structures and processes, or the societal, governmental and economic systems that influence organizational results. These have received much attention. Business schools focus in these areas. Most managers embrace those responsibilities. What needs more attention is how organizational social systems operate and the leader’s role as a part of that system.

Relational communication

The health industry reported that $21B was spent in 2008 on medical errors. At first the healthcare industry responded with checklists and procedures. Then research began in the area of relational communication.

Relational communication leads to compliance

Relational communication(RC) is a subset of interpersonal communication that focuses on the expression and interpretation of messages within close relationships. Healthcare has focused on researchingthis RC. Doctors reported that RC increased medication compliance in patients. Nurses reported that RC improves relationships and compliance. Linda Finch, PhD, reports that patients perceive relational communication as“calm, comfortable, caring, interested, sincere, accepting and respectful.” Finch goes on to say that caring was a new element of RC that was not mentioned by other researchers in this field. Now building a “culture of caring” is the standard in healthcare. In my safety practice I also found that caring was an important factor in both physical and psychological safety. “Management doesn’t care about us,” was the most consistent reason employees gave for lack of trust.

Relational conversations and assessment process

No one needs convincing that trust is a huge factor in organizational performance. What works and we avoid in most cases are the frequent 1-1 conversations between leader and follower. These hold the potential for transformation. I call these relational conversationsas opposed to relational communicationbecause I want to emphasize their back and forth nature. They are also personal in that they are specific to what is going on between two individuals. They hold the intention to build relationships based on mutual respect, trust and benefit. That creates an opening keep the new ways of interacting alive.

Relational Assessment Process

What works and we avoid in most cases are the frequent 1-1 conversations between leader and follower that hold the potential for transformation. These relational conversations with direct reports and others in their organization hold the intention to build relationships based on mutual respect, trust and benefit. Thus, opening an opportunity to sustain new ways of interacting and build collaborative relationships.

I am proposing that we can keep the perception survey for metrics and that we need to incorporate a relational assessment process that includes that employs 1-1 conversations between manager and supervisor and their direct reports. In reality, this process already exists in high performance social systems. It means that we engage in assessing how we are working together and communicating within the container of a psychologically safe manager-employee relationship. Since it is impossible to completely do away with anxiety or miscommunications, doing this on a regular basis provides the opportunity to get the relationship back on track in a timely fashion if needed. 

Adopting this approach would mean moving the bulk of the assessment process to executives, supervisors and managers. The leader would use the data from anonymous surveys to transform relationships through 1-1 or team relational conversations on a frequent basis. The objective is to establish the trust and open communication needed for timely 2-way feedback. In this way assessment is continuous rather than confined to performance evaluations or company wide surveys.

Conversation is the vehicle for change and transformation

The idea of managing people or safety performance via static rules and procedures in a dynamically changing environment has passed us by. Complexity management gurus Ralph Stacey and Patricia Shaw have presented enough evidence to convince me that conversation is the vehicle for change and transformation. As the conversation goes, so does the organization. You can read some of the basics about how it works here.

A continuous relational conversation process is the most effective way to keep a pulse on your organization and collect data. What data would be collected? Managers are pretty good at setting performance metrics or envisioning how the company wants to be perceived by the customer. It is harder for them to use social systems thinking to achieve those goals. Managers may not know what to ask or realize that the stress of conflict; insufficient resources or lack of clarity on priorities can lower their results.

Frequent manager/employee conversations build and maintain trust because it communicates caring and respect. These conversations are most effective when structured around building relationships and organizational priorities.  

Implementing Relational Assessments

At a corporate level we would begin with the executive team. You could also do it with the leadership team of a facility or department. I have seen supervisors be successful when the culture around them is dysfunctional. When the top person in that social system starts the process they can begin to experience the benefits of frequent 1-1 relational conversations with direct reports. In that way they can support their direct reports in doing the same.

To build capability, depending on the level of trust, it is best to begin the conversation process with non-threatening topics like:

  • Tell me about yourself, your work. 
  • What do you like about it?
  • What’s hard?
  • How can I support you?
  • What do you see as the next step in your career?

As trust builds, questions can be added to gather perceptions about the items that are important to measure. This would be determined by the leadership team, which usually includes both HR and a safety professional.

Eventually the conversations spread peer to peer throughout the organization. The transformation is reflected in the way every organizational member interacts, shares information and includes or excludes each other. Unlike the dreaded “performance evaluations” or “feedback” employees look forward to these conversations and see them as contributing to making their work feel more meaningful. Managers may react by saying that they don’t have time to participate. Yet the process, when done right, leads to creating the trust and open communication that is important to high performance. In turn, willing employees take on responsibilities that free up some of the managers’ time.


Unlike the dreaded “performance evaluations” or “feedback” employees look forward to these conversations and see them as contributing to making their work feel more meaningful. 

Engaging in this process keeps leaders perpetually engaged in assessment. This is in alignment with the reality of organizations. Things are constantly changing. There are unexpected challenges, customer demands, changes in the economy and so forth.

A conversational approach to organizational assessment would establish a relationship between boss and direct report that would make shared understanding and accurate communication possible. The boss would have an opportunity to ask questions and check on his or her understanding of the employee’s needs and emotional state. This would lead to making corrective actions more effective.

In short the conversational assessment becomes a satisfying socio-emotional event. Whereas the traditional survey process often reduces psychological safety for managers who feel threatened by negative feedback. Without conversation the manager and employee may not completely understand the other’s perspective. Neither feels heard. Such a process does nothing to increase trust and open communication. 

As I engaged a group of supervisors and managers in digesting an employee engagement survey several of them asked, “When are our needs going to be considered? Everything is always about the employees.” Good point, everyone’s emotional needs have equal importance because the level of organizational performance reflects the way we treat each of its members. 

Rosa Carrillo is available for consultation and key notes.

Leave a Comment

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