Behavior vs. Attitudes: What Do We Address?

Terry E. McSween, Ph.D. and
Grainne A. Matthews, Ph.D.

In spite of what some people might tell you, attitudes are important to a behaviorist. Behavior analysts are concerned with all aspects of a person’s performance, including both what we do and how we do it. The difference between behavioral safety and other approaches to safety improvement lies simply in the difference in how we allocate that most scarce of corporate resources, time. In today’s business environment, companies must deal with many rapidly changing features of the business world with fewer and fewer personnel. The problem of safety is no longer one of know-how but of time-to. So, the question for safety professionals today, more than at any time in the past, is how to apply our time most effectively.

The issue is one of emphasis and getting the greatest impact for your safety dollar. That is, what gives the biggest bang for the buck? Some consultants say that, in their attempts to address behavior, they take a “holistic approach” that addresses attitudes as part of their intervention to improve safety. What could be wrong with taking an eclectic approach that tries to address attitudes by drawing on theories and research from different areas of psychology?

This question is an important one to practitioners who are always wrestling with promoting a safe work place on limited funds. Behavioral psychologists have two problems with trying to address attitudes. The first problem is very simply cost effectiveness. The only way to affect attitudes directly is through education . The educational strategy may be based on logic, emotion, or experiential exercise. We use a logical approach when we provide employees with information about safe practices and the consequences of working unsafely. We arrange for testimonials and we show films to create emotional reactions to motivate people to work safely. Finally, we involve employees in structured activities to provide insights, knowledge, or skills. Such experiential exercises are most often used to teach communication, teamwork, or self-awareness, often discussing the implications of that new knowledge or skills for safety.

All of these educational strategies have utility. The problem is that we have already done so much training. Most organizations today have already made huge investments in training. Often they simply cannot afford to add more training than is absolutely necessary. So one of the primary issues is cost effectiveness.

So if we cannot, or prefer not to add additional education to address attitudes, what do we do? The answer is to treat the culture of the organization as a system, and fix the system so that it promotes safe work practices and discourages unsafe practices. Culture consists of the formal systems of the organization and the social environment of the work place. We have to design the formal systems to promote the kind of social behavior that supports safe work practices. (Culture is addressed from a behavioral perspective in more detail our other paper in this series.) In truth, we may have to do additional training to provide skills that support the new systems, but the focus is on the behaviors necessary to support the process, not on attitudes.

Given that both true behavioral and “holistic” interventions use training, are the differences real or are they just different ways of talking about the same thing? To answer that question requires that we address the second problem with “attitude” and take a much closer look at the basic issue of behavior versus attitude.

What do we mean when we talk about “attitudes”?

For example, what do we mean when we say John has a good safety attitude or that Jane has a bad attitude about safety? We might mean any of three things, but we probably mean a combination of the three.

The first thing we mean when we talk about a safety attitude is what a person says about safety. For example, we might say John has a good attitude toward safety because he often suggests safer ways to complete a task, he frequently participates in discussions during safety meetings, or he knows a lot about safety regulations.

Unfortunately for our safety efforts, the correlation between what we say and what we do is not a reliable one. Each of us can probably think of many things that we tell others and ourselves are important and yet we might rarely do those things. Exercise is a classic example.

The second thing we mean when we talk about a safety attitude is what a person does with regard to safety. We might say that Jane has a bad safety attitude because we observe that she rarely follows safety practices, she never shows new workers how to complete a job safely, or she frequently avoids safety meeting.

Here we are talking about the probability of certain types of behaviors occurring. A good safety attitude means a high probability that someone will engage in some of the many behaviors that reduce risk and a bad safety attitude means we expect few of those behaviors to occur.

There is a third possibility when we talk about attitudes. We might mean that safe behavior is caused by a safe attitude. In this case, we are falling into the trap of turning a description into an explanation. This is an example of an explanatory fiction.

It is often quite useful to identify and label relatively stable characteristics of a person’s behavior. It allows us to prepare for the anticipated behavior. For example, if someone tells you that your new boss has a positive attitude about behavioral safety, you can expect that she will reward you for implementing behavioral safety and will support you when you propose expanding the process to other sites.

However, a problem arises when we use the name of a collection of behaviors as an explanation for those behaviors. For example, if you ask an employee’s supervisor why the employee doesn’t wear her PPE and the supervisor tells you that it’s because the employee has a bad attitude toward safety, the attitude is no longer a description, but has become an explanation. When the only evidence for the bad safety attitude is the same behavior that the bad safety attitude explains, we have fallen into a trap of circular reasoning. We have created an explanatory fiction.

Such fictions are actually harmful because they distract us from the true causes of observed behavior - the systems and practices that make up the organization. They imply that we understand the behavior when we might not. They short-circuit a careful analysis of the real reasons the worker is not wearing her PPE.

An understanding of the science of behavior is important for today’s business leaders and safety professionals alike. Change is occurring so fast that managers must have a logical and reliable framework within which to organize and make sense of new information and new technology. Science provides the only coherent system that can help us to sort through the claims and counterclaims of management gurus, training celebrities, and charismatic consultants. Without a scientific understanding of behavior, we are likely to adopt strategies and tactics that are not only ineffective but also actually counterproductive. We are all familiar with the fallout within an organization when another program fails or a former favorite innovation is pushed aside. We lose credibility and weaken morale each time this occurs. The only way to avoid this is to possess and use a scientific approach.

If we explain our hypothetical worker’s behavior by referring to an internal attitude, we have also led ourselves into more than just logical difficulties. B.F. Skinner, the founder of the scientific study of behavior, maintained that this kind of reasoning contributed to our inability to solve urgent social problems, such as safety in the work place. If the cause of unsafe behavior is internal attitudes, then the solution must lie in changing the world inside our heads. If, on the other hand, unsafe behavior is explained by external causes, such as social pressure to break the rules, or management pressure to get the product out the door no matter the cost, then we have a chance. Then we have an opportunity to change the system so that it supports safe behavior instead of encouraging unsafe work habits.

What are attitudes?

From a behavioral perspective, attitudes are statements that typically describe our behavior in certain situations. They may also describe our emotional response to, or how we feel about a given situation. To oversimplify the situation, we might say that there are two different kinds of behavior, what we say about how we act in a given situation, and how we actually act in that given situation. One of the problems is that while these two different types of behavior are often correlated, it is not a direct correlation. In other words, people do not always act the way they say they will act. That is the crux of the problem. It is often easy to get people to say they will behave in a given way, but then they leave the classroom and go back to their old habits. For example, with a well designed workshop, we can get most participants to discuss the importance of wearing seatbelts, and even to say that they will wear the seat belts. The problem is that once they leave the classroom, only about 65% of them will actually wear their seat belts, although it is an act required by law in most states. So the point is that even if you address attitudes successfully, as measured at the end an educational process, you still have to address the behavior on the job – 65% compliance isn’t good enough for those of us working in industry.

Moreover, if, in the end, you have to address the behavior anyway, why spend the additional resources to address attitudes. The research literature on behavioral safety shows us how to create a social environment that promotes safe work practices and achieve statistically significant reductions in injury accidents. There is no research published in refereed journals that suggest that a “holistic approach” focusing on attitudes adds anything to the success of these efforts. The implications are clear, we need to focus on behavior, and we need to provide the social skills necessary for such an approach to be effective. A “holistic approach” may sound good, but scientific research has yet to show that its elements add value.

Published in Proceedings of the 1998 ASSE Behavioral Safety Symposium: Light Up Safety in the New Millenium, American Society of Safety Engineers, Orlando, FL, 1998, p. 149-154.