Observe Current Levels of the Target Behavior

One mistake supervisors often make is to begin intervening before they have adequately observed the target behavior as it is. Once we have become aware of a problem, it is very difficult to withhold instituting a change, but careful observation is an essential step in a successful behavior modification program. In addition to making a specific definition of the target behavior, you must analyze how the behavior is controlled and then establish a baseline from which to measure the success of the intervention. These three tasks can often be performed simultaneously, but for clarity of discussion they are presented here as discrete steps.

Discovering Consequences

Identifying consequences of the target behavior can give an important clue to how to modify that behavior. Generally, behaviors that occur often-be they desirable or undesirable-are maintained by positive consequences, and behaviors that occur infrequently are maintained (or suppressed) by negative consequences. Try to think of positive consequences as "turning on" something good {positive reinforcement) or "turning off" something bad {negative reinforcement), and negative consequences as turning off something good {positive punishment or extinction) or turning on something bad {negative punishment).



Because consequences have such a heavy influence on how often a behavior will occur in the future, identifying them becomes the next step in specifying the problem. The best way to identify a behavior-consequence pattern is through observation. Note and write down what happens after the behavior: Is there anything that seems to be turned on or off after the behavior? Is there a pattern or recurring event? Is there more than one consequence? When Georgia observed what happened after Otto made a negative comment, she discovered a curious thing: Her own behavior was a consequence-she consistently responded with defensive comments. If we hypothesize that Otto perceives Georgia's defensiveness as positive, we can say Otto's negative behavior is maintained by the positive reinforcer of Georgia's defensive commenting. Often there is more than one consequence: Georgia found that Otto's negative comments were often followed by positive attention from the other salespeople (such as chuckling and knowing looks). The attention from others was probably a more powerful reinforcer than was Georgia's defensiveness. Had she stopped observing as soon as she had identified one consequence, she might have failed in her change program. See if you can identify the consequences in the following situations:

1. A claims adjuster trainee handled five claims. Three claims had missing information, and two claims were complete. The super visor commented on the thoroughness of the two completed claims.
   
   
2. The previous supervisor of the typing pool allowed those typists who completed their assignments rapidly to take discretionary time. The new supervisor stopped the practice, and insisted that all typists look productive at all times.
   
   
3. A new employee completed the first assignment promptly by the deadline, and handed it to the supervisor. The supervisor made no comment.
   
   
4. William submitted to the planning committee a proposal that contained several cost-effective changes. In William's presence, his supervisor spoke to the department director and made positive comments about the proposal.
   
   
5. Bob spent several evenings and most of the weekend putting together an outline for a new sales promotion. His supervisor responded by pointing out three major weak points.
   
   
6. Alice completed the quarterly report a day early and gave it to her supervisor. The supervisor gave her an additional assignment to complete during the extra time she created.
   

Now compare the consequences you identified with those I identified.

1. Completing claims was reinforced by a positive comment. Not completing claims was being extinguished by the absence of a response.
   
   
2. Rapid typing was punished by the new supervisor. The supervisor has turned off something good: discretionary time.
   
   
3. The employee's on-time behavior was being extinguished by the supervisor's lack of response.
   
   
4. William's suggestions were reinforced.
   
   
5. Bob's conscientious work was punished.
   
   
6. Alice's rapid work was punished, unless Alice perceives extra assignments as positive, that is, as an opportunity to get ahead.
   

The easiest kind of intervention is one in which you alter the consequences. But in order to do that, you have to be able to provide a new consequence for the target behavior. Unfortunately, it isn't always possible to separate the reinforcement from the behavior-this happens when behaviors consume the reinforcer. For example, smoking is reinforced by inhaling the smoke; overeating is reinforced by ingestions of food; too many drinks at lunch, by the drinks themselves; talking on the phone, by the verbal responses. With such consummatory behaviors you can easily identify the consequence, but rarely change it. Thus when the target behavior is consummatory you will need to look to the antecedent for a clue as to how to modify that behavior.

Identification of the consequences can also become obscured by intermittent reinforcement and avoidance behaviors. In these cases, understanding antecedents will also be essential to understanding how the target behavior functions. Some behaviors are resistant to extinction because they are maintained on an intermittent schedule of reinforcement. In other words, the behavior goes unreinforced many times, but will continue as long as there is an occasional reinforcement. Intermittent reinforcers are difficult to identify. If the target behavior is reinforced only one time in ten, you would have to observe the behavior 30 times just to obtain three examples of the consequence-and reliable conclusions cannot be drawn from only three examples. Obviously, such lengthy observation is not generally feasible.

The consequences of avoidance behavior are even more difficult to identify. An avoidance behavior is performed to avoid some punishing consequence. If it is successful in avoiding punishment {positive consequence), the avoidance behavior will be more likely to occur again in the same situation (antecedent). Obviously, if the punishing consequence does not occur, it cannot be observed. Sometimes, by imagining what would happen if a person were to perform differently in a certain situation, we can guess at the consequence that is being avoided. If one of the possible consequences turns out to be punishing, this suggests that the target behavior is an avoidance behavior.

For example, when the radio station manager asked Charles what would happen if he were to ask Rosemary directly for the projection, Charles said, "She'd just ask a bunch of questions and make me de fend every decision I made!" Because Charles stated this in a way that suggested it was punishing to him, his manager could suspect that Charles's procrastination was tied to his avoidance of directing Rosemary.

Discovering Antecedents

When a stimulus is presented each time a behavior is reinforced, that antecedent stimulus eventually gains the power to evoke the behavior. Likewise, when a behavior is repeatedly punished in the presence of a particular stimulus, that stimulus soon becomes associated with punishment and signals to the person that punishment is imminent. If the person avoids punishment, the stimulus can become an antecedent that signals future avoidance behavior. Each time the person avoids the punishment, the antecedent becomes more powerful in its ability to evoke avoidance behavior. Because antecedents can become very powerful in turning behaviors on and off, it is important that these cues be identified.

To discover antecedents, the events that occur just before the target behavior must be identified and recorded. Once several examples have been collected, a common theme can usually be identified. Consider Melody, who makes personal calls during work time. Through observation, her supervisor discovered that Melody frequently made a phone call after she finished typing an assignment. Here the antecedent that evoked the consummatory behavior of making phone calls was probably a finished assignment.

When you have difficulty identifying what occurs just before the target behavior, a complete description of the situation can often reveal the antecedent. When Georgia wrote a complete description of the situation in which Otto's problem behavior occurred, she discovered the antecedent: "Whenever I mention my innovative program in the weekly staff meeting, Otto makes negative comments about the program." This description suggests that the antecedent of Otto's negative commenting is "mention of the innovative program." And indeed, when Georgia carefully observed and recorded each instance of Otto's negative comments in one staff meeting, she confirmed this: Each mention of the innovative program was followed by a negative comment by Otto. Here is Georgia's functional analysis of Otto's behavior.

A.    A Antecedent Mentioned new responsibilities in innovative program Discussed sales promotion under innovative program Discussed procedures for keeping records

B.    B Behavior Quo questioned necessity of the innovative program Otto said promotion wouldn't work Otto said old method was a lot better and cheaper 

C.    C Consequence Told Otto why innovative program is necessary Told Otto I had a lot of experience Told Otto that I am the manager in innovative program

Establishing the Baseline

One way in which behavior modification differs from other approaches to managing people is the baseline. The baseline tells how often the target behavior is occurring now-before intervention. It is a method of accountability and evaluation. The baseline helps to monitor the program and rapidly detect an ineffective intervention.

Consequently, it is cost-effective, and is also an essential step in a successful behavior change program.

Making the functional analysis and gathering baseline data frequently overlap, because both require careful observation. However, the type of observation differs. Antecedents and consequences are identified by recording descriptions of events. To establish the base line, count how many times the behavior occurs and record that number,

To obtain an accurate and reliable baseline, the target behavior must be clearly specified. If your description of the target behavior is vague or ambiguous, you will find out quickly. Gathering baseline data involves counting each occurrence of the behavior-in-situation.

For example, Georgia counted each time Otto made a negative comment in the staff meeting about the innovative program. At the end of the meeting, she recorded the number of negative comments on a graph. In this way she could determine the frequency of Otto's negative comments over time.

Sometimes the concern with the target behavior is not a matter of how often it occurs but the duration of the occurrence. Suppose the problem with Melody's personal calls was not that she made them but that she talked for an average of 15 minutes. Here, her supervisor would record the length of the calls. Generally, count either the number of separate times people perform the behavior or the amount of time they spend performing the behavior.