I always make mistakes and take calculated risks on the tennis court. When my focus or concentration wavers, I unintentionally miss-hit the ball or fail to get in position for a solid return. Sometimes I'll rush the net when I shouldn't, or hit the ball long when trying to land a baseline corner.

What I should do is quickly re-focus my attention or reconsider the risks I take. Instead, I often yell at myself. Occasionally I'll slam the ball against the fence or toss my racket. Does this self-critical "punishment" ever help? Of course not. It only makes matters worse. The same is true for your golf game, and for meeting the continuous challenge of preventing injuries in the workplace.

Still, we rely too much on punishment to correct behavior. In this article, I want to show the futility of this approach. Errors, or cognitive failures and mistakes (that result from our mental processes involving awareness and judgment), are unintentional and often caused by environmental factors. And when errors are intentional (as in calculated risks), the person didn't intend to cause an injury. Rather there were factors that influenced the decision to take the risk. The reasons for errors and calculated risks need to be discovered and addressed.

Before meting out punishment, answer the following seven questions. In most cases, you'll find another type of corrective action is more appropriate.

Was a specific rule or regulation violated? If you answer "no," punishment is obviously unfair. Does this mean you need to write more rules or document more regulations? I don't think so. You can't write a rule for every possible unsafe behavior. And since human errors are unintentional, rules won't decrease them. We need to allow for the possibility that non-compliance with a rule or regulation can be unintended. This leads to the next question.

Was the behavior intentional? All human error is unintentional. We mean well but have cognitive failures or "brain cramps." Psychologists call these errors slips or lapses and they are typically due to limitations of attention, memory, or information processing.

Have you ever walked into a room and forgot why you were there? Locked yourself out of your car with the keys still inside? Left your house only to return to get something you forgot?

In the workplace, research has shown these types of errors increase with experience on the job. Skilled people put their actions on "automatic pilot," and perhaps add other behaviors to the situation. How many of us fiddle with a cassette tape or juggle a cellular phone while driving? You can see how an error can easily occur. This "unconscious incompetence" needs to be corrected, but certainly not through punishment.

There is also "conscious incompetence." Sometimes poor judgment is used to intentionally take a risk. This conscious behavior can be classified as either a mistake or a calculated risk. Have you ever miscalculated a parking space and scraped an adjacent vehicle? Pressed the brake pedal too quickly on a slippery road or pumped the brakes in an antilock system? Parking and braking are frequent and intentional driving behaviors, but under the particular circumstances they are mistakes.

Now suppose you don't buckle your safety belt. You know this behavior is unsafe, but you decide to take a calculated risk. In this case, unlike a mistake, you're aware that your behavior is inappropriate. The deliberate or willful aspect of calculated risks might seem to warrant punishment, but punishment won't convince people that their judgment was defective. And that's what is needed to change conscious incompetence to conscious competence.

Was a rule knowingly violated? Researchers have proposed that the more knowledge or skill we have at doing something, the less likely we are to demonstrate poor judgment. So the driving mistakes listed above occur more often from inexperience or poor training. On the other hand, the tendency to take a calculated risk increases with experience on the job. This is human nature, and it won't be changed with punishment.

Some errors occur because the rule or proper safe behavior was not known. And it's possible for an experienced worker to forget or inadvertently overlook a rule. Training and behavior-based observation and feedback can reduce these types of errors, but punishment certainly won't help.

How much were other employees endangered? Failing to lock-out a power source during repair work is the behavior most often targeted for punishment. This is in line with what I've heard many safety professionals say: They only punish employees when a particular behavior risks a severe injury or fatality, or places many individuals in danger. Some pros are quick to add that these risks must involve willful intentions and prior knowledge.

Now if an employee willingly and knowingly avoided a lock-out procedure to put himself and others at risk, then the severest punishment is relevant. Actually, this person should be fired immediately. But this rationale for risky behavior is very rare.

Many dangerous behaviors are mistakes resulting from poor judgment, not an unconscious or conscious desire to circumvent policies and hurt someone. And when a calculated risk is taken, it's not with the idea that someone will get hurt. Often specific characteristics of the work environment or culture enable or even encourage a calculated risk.

What supports the behavior? This is the most important question of all. People don't make errors or take calculated risks in a vacuum. Poor judgment occurs for a reason. And it's important to learn an employee's rationale for taking a risk. This leads to truly useful corrective action.

Did the employee lack knowledge or skills? Was a demanding supervisor or peer pressure involved? Did equipment design invite error with poorly labeled controls? Was the "safe way" inconvenient, uncomfortable, or cumbersome?

Let's look at the organizational culture: Is safety taken seriously only after an injury? Is safety performance judged only in terms of injuries reported per month, instead of the number of preventive activities? These are only some of the questions that need to be asked. Punishment will only make answers-and a practical corrective action plan-harder to come by.

How frequent is the behavior? How many at-risk behaviors typically occur before leading to an injury? H. W. Heinrich estimated 300 near-misses per one major injury, and Frank Bird observed this ratio to be 600 to one. Both Heinrich and Bird presumed numerous at-risk behaviors occur before even a near-miss is experienced, let alone an injury.

So what good is it to punish one of many risky behaviors? If the behavior is an error, punishment will only stifle reporting and potential remedies. If the probability of getting caught while taking a calculated risk is low-and it's minuscule if you wait until an injury occurs-any threat of being punished will have little behavioral impact. Remember, punishment reduces involvement in safety improvement efforts more than it reduces risks.

How often have others escaped punishment? To punish an employee for a behavior that others have performed without receiving similar punishment is one sure way to lose credibility and turn a person against your safety efforts.

To summarize: If the threat of personal injury is not sufficient to motivate consistent safe behavior, it doesn't help to add one more threat (punishment) to the situation. We need open and frank discussions with the people working at risk to analyze and change management practices, equipment, or organizational systems that contribute to much of the at-risk behavior we see in the workplace. This is only possible when the threat of punishment is removed.

Turn to the "Reader Response" department on page 4 for other opinions about the use of punishment.