The Anatomy of Pencil Whipping

The term pencil whipping is a euphemism used to describe when workers, supervisors and safety managers fill out safety forms such as behavior-based safety observation cards, without actually conducting the observation or audit. It's time to point to the elephant in the room and acknowledge pencil whipping within our safety programs. This article seeks to explain and provide solutions for the environment that causes pencil whipping.

Forms, Forms, Forms

I looked at a supervisor's office at a major construction site. It was more overloaded than mine (yet much neater). The desk was full of paper, forms upon forms to fill out. These included operational forms, human resources forms, quality checks, surveys from upper management and, of course, a plethora of safety forms: job safety analysis forms to be completed every shift, first-aid forms, safety audits, injury investigation reports, safety observations, safety work orders, safety suggestion forms and process safety checks. I asked the supervisor when he had time to go out to the site. "And give up my desk job?" he replied jokingly.

Employees are increasingly asked to complete forms. Permits, equipment inspections, work orders and behavioral observations are just a few that focus on safety.

All of these forms are designed to reduce injuries. A well-designed behavior observation card, for example, not only provides the script for reinforcing feedback, it also can identify risk and give the observer an opportunity to comment on the cause of the actions and suggest fixes (Geller, 1996; McSween, 1995; Sulzer-Azaroff & Austin, 2000).

These forms serve an important purpose. As research and real-world application shows, the more of them that are completed, the more injuries are reduced (Myers, McSween, Medina, et al., 2010). Figure 1 (p. 48) shows that increasing annual participation in peerto-peer observations was associated with decreases in OSHA-recordable injuries in the context of a behaviorbased safety program at a petrochemical refinery (Ludwig & Austin, 2012).

This is because more observations result in more one-on-one feedback, which is the most powerful (and cheapest) motivator of safe behavior. As an added benefit, completed observation cards can be analyzed and trended to identify the most prominent areas of risk. When these areas are targeted for improvement, equipment is updated, protection is worn and shortcuts are no longer needed. This reduces the threat of injury as well (Cooper, 2006; Geller, 1996; McSween, 1995).

That is, if the forms are actually used for this purpose. Enter pencil whip.

Pencil whip is a verb, which means it is an action, something someone does, and does intentionally. The term is so prolific that it made it into the Wiktionary: "To complete a form, record or document without having performed the implied work or without supporting data or evidence." In Australia's vast mining industry it is called "tic and flick." The idea is that you are filling out forms so fast making up the data that the end of your pencil is whipping in the air.

How Do You Know Data Are Getting Pencil Whipped?

Some clues to help you know whether data are getting pencil whipped: Look at behavioral observation data for big jumps in cards being submitted in just a couple of days. Look at those cards and you will find multiple cards turned in at the same time. The same person would complete these cards in just a couple of days (some may use different pens and slightly alter the handwriting). The person certainly would not take the time to write in comments or suggestions (Dagen, Alavosius & Harshbarger, in press).

You will also notice that at-risk behavior is rarely reported on such cards. Indeed, if the at-risk behavior percentage is below 2% on nearly all behaviors, you have a pencil whip problem.

Perhaps the easiest way to find out about pencil whipping is just to ask. Most will admit to some pencil whipping but typically you'll hear "supervisors tell us to do observations because we need to make our numbers," or "I do the card afterward from memory," or "I do the cards to get in the drawing for the gift cards. I nearly forgot before the deadline." Or, simply ask yourself. Admit it, you already know. In fact, you may be enabling pencil whipping because having a lot of observations makes you look good.

The Case of the 1,000 Observations a Month

A mining construction general contractor had the best intentions when the experts told him to monitor his company's behavior observation process measures. He was coached to ask about the number of observations his contracted workforce was performing. When I met with him after the process was in place for more than a year, he was convinced that behavior observations did not work. I would have drawn the same conclusion.

"We are getting more than 1,000 observation cards turned in a month, yet we keep getting around two incidents each month." He concluded, "It doesn't work in construction." He described a number of times when his safety people reviewed cards from the previous day and saw that an action was performed safely 100% of the time.

"Walking under loads was a big one, it's a real risk on a construction site with cranes everywhere. Well, dozens of cards were turned in indicating that no one was walking under these cranes, 100% safe." His teeth clinched a bit, "The day these cards were turned in, my safety guy went out to that work site and saw nearly all the workers walking under a crane. There were no barricades or markings. In fact, that pathway was the only route between two work areas. There is no way these cards are accurate. The employees only want to make each other look good."

It was time to ask questions. "Confidentially, tell me what's going on."

Employees: "I never heard of the program. I've not been observed."

Then I started asking questions up the line.

General contractor: "I did what I was coached. I asked to see the contractor participation numbers."

Construction manager (reporting to the general contractor): "I told the contractors to get me their behavior observation numbers."

Contractor manager: "The general contractor wants the numbers, I pressed my supervisors to get those observations done so we looked good. I don't want to be the one who loses this contract over a safety program."

Supervisors (when presented with the dozens of cards in their handwriting): "I filled out the cards for the employees ... at the end of the shift."

Translation: Pencil whip.

Production is a numbers game, and it should be. But when "getting the numbers" is applied to safety, well, you get the numbers.

Going "All-In" With Incentives

Another well-intentioned tactic to encourage observations is to offer incentives. I've seen many increasingly innovative incentives. Raffles seem to be the most popular because they are more like a game than an incentive.

A contractor organization at a refinery had developed a popular game similar to Texas Hold 'Em to promote observations. It was pretty cool. Each month, employees received a playing card for every five observations they turned in (up to four playing cards). An employee could get two additional cards for selfobservations, two cards for completing observation training modules, and two more for special safety programming such as participating in inspections. The employee who had the best poker hand won a $50 gift card.

This game generated a lot of observations, but they came in waves right before the end of the month. The company has a great behavioral safety program and its injury rate is near zero-amazing for a construction company. But examining the cards completed in the 4 days leading up to the end of the month reveals the telltale signs of pencil whipping (one person turning in multiple cards a day, all 100% safe, no comments). Minimal indications of at-risk behavior could demonstrate fake reporting (also known as pencil whipping) by the observer (Cooper, 2006).

Similarly, zero-injury incentives may reduce reporting behaviors if the reporting of an incident would void the reward for a worker and/or peers (Agnew & Daniels, 2010; McSween, 1995). Both incentive and disincentive programs can actually discourage employees from reporting injuries or incidents such as close calls or Witnessing others' at-risk behavior (Geller, 1996; Geller, 2005; McSween, 1995). You get what you pay for.

How Pencil Whipping Evolves

The following is based on a study by Boitnott and Ludwig (2012).

A company operating a fleet of ocean research vessels developed a behavioral safety program. It was replete with common observation cards, online observation card recording software, electronic training, local steering teams and an excellent data analysis program. Mariners would retrieve a card, perform their observations, give feedback, then enter observation data themselves on an on-board computer connected to the company intranet.

Some vessels had better safety records than others and some had more observations recorded than others. In an effort to increase observations on some vessels, one of my students helped this company devise a shorter observation form. The original form had 18 items to observe and score as safe or at-risk. He reviewed the incident and first-aid files to find the eight behaviors most likely to result in an injury. The company then adapted the cards to include only those eight behaviors and put the cards on the decks of three vessels, and told workers that the cards were easier and quicker to complete.

It didn't work; the observation counts did not increase. However, an interesting thing happened: employees kept reporting on the old behaviors even though the new cards dropped 10 of the behaviors and listed only eight. The computer software used still listed all 18 items, so employees reported on all 18 items, not just the eight on the card.

In our interviews afterward we were told, "Well, the guys have used the cards for so long they just rely on their memory to enter the observations." They were no longer using cards at all.

It occurred to me that this may be how pencil whipping evolves. At first, employees use the card and do a real observation of a peer. Then the employee likely provides the peer feedback and the data the employee turns in is accurate.

When employees memorize the cards, it stands to reason that observations happen rather quickly. Feedback is less likely to happen over time. When the data are finally recorded on a card or, in this case, in the computer, accuracy goes down. With memory comes forgetting. We know that only seven (±2) pieces of information can stay in short-term memory at any one time and those memories only last a couple minutes at best.

Because we are likely to forget, we tend to rely on our overall impression of the person, the type of work being performed or how safe things tend to be. This is called the halo effect, not because we see the person as angelic, but because we let our overall impression of the person or situation help us guess (Rosenzweig, 2009). Indeed, guessing further reduces accuracy and we miss the opportunity to identify risks.

Let's follow the trend. Employees or supervisors now have experience not using the card yet still getting credit for turning in an observation. They figure out quickly that the only thing they really get credit for is turning in a completed card or, in this case, entering data. This can be done, they leam, without even doing an observation, much less giving feedback.

The pencil whip is in good shape and the observation program is in jeopardy. Worse, people's lives are more in danger.

Pencil whipping occurs because it is the path of least resistance. Don't blame the whipper. In fact, the observation form process is perfectly designed to produce pencil whipping. Redesign it to reinforce the conversations it is designed to facilitate.

Suggestions to Temper Pencil Whipping

Following are several ways to help minimize pencil whipping.

1) Name the demon. In the Bible, Jesus gets rid of demons by calling out their names. There are parallels in most religions and tribal customs. The wisdom rings true here. Bring pencil whipping out of the closed offices or workstations and into the light. Admit it is a problem by name. Don't accuse, instead inquire: "How can we change this?"

2) Drop incentives and quotas in favor of rewards. Incentives are planned and advertised as a positive consequence of action. Rewards, on the other hand, are seemingly spontaneous acts of gratitude (Geller, 1996). They don't have to be big, but they must be fair, and they must be based on the outcomes of behaviors that employees can control. They should specify the behaviors that happened and offer genuine appreciation or recognition.

In the case of behavior observations, rewards should be given for identification of at-risk behaviors and comments, especially those that lead to fixes.

3) Provide group feedback abundantly. Make visible changes. Reinforce the notion that making an accurate observation, giving feedback and reporting all make a difference. Doing this shows employees that life is better because of their actions: the workplace is safer and things are changing. Provide the group feedback about their success: rising observation counts, higher participation, injury reduction and, most of all, at-risk behaviors identified (McSween, 1995).

A contractor fitting insulation at a refinery has a behavioral observation program effective at identifying risk. One at-risk behavior that kept showing up was working outside in hazardous weather. The work team was thanked for providing the data to show that this was an unnecessary risk that workers felt they had to take. Management then instituted a clear policy statement on work procedures in inclement weather, including stopping work.

After employees experienced this victory, they began reporting at-risk behaviors around the nonuse of respirators. This behavior was quickly fixed by changing an uncomfortable chinstrap. In both cases, these behaviors were now performed safely in 100% of reported cases. Also, employees were more likely to identify other, more personal areas of risk such as body mechanics. This reporting behavior is the opposite of pencil whipping.

4) Change the behavior observation card frequently. Sometimes the problem is as simple as the old card has gotten stale. Fight the impulse to use the same card across all work units forever. The impulse comes from the great data analysis you can get from long-term common observations (I fall prey to it myself as a researcher).

Shortening the card could reduce cost to participate in the program thus increasing the likelihood of behavior. Geller (1996) argues that longer versions may be overwhelming for employees, thus discouraging participation. McSween (1995) advises shortening by targeting only behaviors that directly correlate with safety outcomes, occur frequently and are easily observed.

One national grocery distribution company designs changes into its behavioral observation process (Ludwig & Harshbarger, 2010). In that process, the employee team chooses only three or four items to appear on a card. Observations and resulting feedback are easy. The data are trustworthy. The percent safe for each behavior is posted prominently and is mentioned at the beginning of each shift. Once one of the behaviors reaches 98% safe for 3 consecutive months, a celebration ensues and the behavior is retired. The team then selects a new behavior to target, the card changes and the system begins anew. Pencil whipping never factors into the equation.

Conclusion

Pencil whipping blinds us to real workplace hazards and the risks being taken in these situations. The suggestions to temper pencil whipping must be customized for different industries as well as individual safety cultures. No solution is 100% foolproof, but the first step is to leam how much pencil whipping is happening at the site or company. This can be done through anonymous surveys and document inspection. Savvy investigators will assess the forms' reliability by spot-checking information that appears on the forms. If the forms indicate that no one is working under cranes, go out and look.

Safety managers may also be well served to review the many forms associated with the safety systems under their control and drop unnecessary ones. Documentation has a way of growing out of control over the years as people add more forms and fields in forms. Fewer forms leads to less pencil whipping.

Pencil whipping cannot be ignored; it must be talked about frankly and monitored. If it is not labeled what it is, then we are all complied in its abundance.

Pencil whip (v): To complete a form, record or document without having performed the implied work.

Vantage Point articles in Professional Safety provide a forum for authors with distinct viewpoints to share their ideas and opinions with ASSE members and the SH&E community. The goal is to encourage and stimulate critical thinking, discussion and debate on matters of concern to the SH&E profession. The views and opinions expressed are strictly those of the author(s) and are not necessarily endorsed by Professional Safety, nor should they be considered an expression of official policy by ASSE.

References

Agnew, J. & Daniels, A.C. (2010). Safe by accident: Take the luck out of safety: Leadership practices that build a sustainable safety culture. Atlanta, GA: Performance Management.

Boitnott, M.K. & Ludwig, T.D. (2012). The cost of participation: Reducing response effort to increase participation and quality in peer-to-peer observations. Paper presented at the conference of Association for Behavior Analysis International, Seattle, WA.

Cooper, D.M. (2006). Exploratory analyses of the effects of managerial support and feedback consequences on behavioral safety maintenance. Journal of Organization Behavior Management, 26,1-41.

Dagen, J.C., Alavosius, M.P. & Harshbarger, D. (in press). Using managerial feedback to improve safety in the petroleum industry. Journal of Organizational Behavior Management.

Geller, E.S. (1996). The psychology of safety. Randor, PA: Chilton Book Co.

Geller, E.S. (2005). Behavior-based safety and occupational risk management. Behavior Modification, 29,539-561.

Ludwig,T.D. & Austin, J.A. (2012). Turning ownership over to the employees: A management "practice" that yielded results. Paper presented at the conference of Association for Behavior Analysis International, Seattle, WA.

Ludwig, T.D. & Harshbarger, D. (2010). Back from the brink: Using behavioral-based safety to management safety after a three-fold increase in production. Paper presented at the conference of Association for Behavioral Analysis International, Austin, TX.

Ludwig, T.D., Harshbarger, D. & Hopkins, B.L. (2010). Sustaining world-class safety performance through innovation. Paper presented at the conference of Association for Behavior Analysis International, Austin, TX.

McSween, T.E. (1995). The values-based safety process: Improving your safety culture -with a behavioral approach. New York, NY: Wiley.

Myers, W.V., McSween, T.E., Medina, R.E., et al. (2010). The implementation and maintenance of a behavioral safety process in a petroleum refinery. Journal of Organization Behavior Management, 30,285-307.

Rosenzweig, P. (2009). The halo effect and the eight other business delusions that deceive managers. New York, NY: Simon and Schuster.

Sulzer-Azaroff, B. & Austin, J. (2000, luly). Does BBS work? Behavior-based safety and injury reduction: A survey of the evidence. Professional Safety, 45(7), 19-24.

Timothy D. Ludwig, Ph.D., is a senior consultant with Safety Performance Solutions. He is also a commissionerfor behavioral safety accreditation at the nonprofit Cambridge Centerfor Behavioral Studies and teaches behavioral psychology at Appalachian State University in Boone, NC. He holds a Ph.D. in Behavioral Psychology from Virginia Tech. Read more at http://safety.doc.com.