Positive results can help you persuade top management to provide funds and support for implementing your program on a broader basis. In addition, program evaluation provides concrete evidence of your managerial skills. Even when results are not as positive as you might have wanted them to be, merely having conducted a systematic evaluation demonstrates that you have essential managerial skills, such as planning and decision making.
Publishing the results of your evaluations in in-house publications and trade journals will give others the opportunity to profit from your activities. Colleagues within your company can make use of tactics that you found effective. In the long run, this will help your company because others will not have to rediscover what you have discovered.
The final reason for careful evaluation has to do with recognizing the individuality of your employees. Management strategies presented or in seminars and workshops are general prescriptions and provide general guidelines. But each employee is an individual who has a unique learning history.
Thus, change programs that were effective with loggers or bakery workers may not be effective with your employees. Likewise, an approach that works with employees in general may not be effective with a particular problem employee. At this time there are a limited number of controlled studies documenting the effectiveness of behavior change programs with individual employees. There is much work to be done in this area, and there are a number of popular organizational publications that are interested in such material. In this section I will briefly describe several common evaluation designs.
Evaluation involves comparing the outcome of your program with a standard. Without a standard or control, objective evaluation would not be possible. Evaluation designs are models for structuring programs, and each design has some kind of built-in comparison. Comparisons can be made within the target group (intragroup) or between the target and a control group. Intragroup comparison means that the behavior of a target person or group is compared with itself: The target's own behavior prior to the introduction of the change program provides the standard for comparison. Here are descriptions of three kinds of intragroup evaluation designs.
The baseline design (AB):
This is the basic evaluation design. Data are collected on the frequency of the target behavior before any change is introduced. Pre-change data are called baseline, and collecting baseline data is continued until a pattern is revealed. The behavior change program is then introduced and data collection is continued. Evaluation consists of comparing the frequency of the target behavior during the baseline period with that same behavior during intervention. The problem with this design is that when a change is indicated it cannot be attributed unequivocally to the intervention or behavior change program; extraneous factors in the environment may have influenced the target behavior. Figure 15 illustrates results from a typical baseline design.
The reversal design (ABA)
This design is similar to the baseline design, except that after the change program has been in effect for a designated period of time, it is withdrawn or reversed. During the reversal phase, if the frequency of the target behavior returns to baseline, then we can put more confidence in the effectiveness of the change program. The incentive program to reduce tardiness in the Mexican factory is an example of this design. The Hermann team reversed the intervention twice, and following each reversal the frequency of the tardiness increased, whereas when the incentive program was in effect it decreased. Figure 16 illustrates results from a typical reversal design.
The staggered design (AB1B2)
The staggered design can be used when two or more behaviors are the target of the intervention. After baseline has been established, the change program is introduced on a staggered schedule. The program with the grocery store clerks illustrates a staggered design. There were three variables targeted: the location of clerks, the speed of customer assistance, and the quantity of merchandise on the shelves. Interventions to change each of these behaviors were introduced one at a time. The data revealed that each behavior changed only after the intervention targeting that behavior was introduced.
Intergroup comparison refers to an evaluation comparison between two different groups. Baseline data are collected on both groups (or individuals). The intervention is introduced with one group but not with the other, and the group that does not participate in the change program is called the control. This kind of design can-and, whenever possible, should-be combined with any of the intra-group designs. The purpose of the second group is to provide a control for factors in the environment, which might affect the behavior of the target subject. The rule of thumb is that there should be only one difference between control and targets, and this difference should be the change program. Control employees should be from the same population-similar in age, sex, race, job title, and so forth.
Controls should continue in the normal workaday routine, should not know they are serving as controls, and should be treated no differently from the way they would be were they not being evaluated.
An exception to this rule is a situation in which there is reason to believe that a noncritical aspect of the change program may affect behavior. An example is special attention. Research has demonstrated that special attention can influence productivity. Thus, productivity might increase in the target group as a result of attention-not of the behavior change program. Without a control for attention, evaluators might wrongly conclude that the behavior change program caused the change.
One way to solve this problem is to expose the control group to special attention, but not to the conditions of the change program. Then, if the results indicate a difference between target and control, you can attach more weight to the conclusion that the change program was instrumental in causing the change. The program with the loggers conducted by Latham and Yukl illustrates this procedure: The controls were merely urged to do their best, whereas the target groups either participated in goal-setting or were assigned a goal. In this way, the control group was exposed to attention and promptings for increased performance, but not to the goal-setting.
When planning to introduce a change program in two departments or with two employees, you might consider using a staggered schedule in which a person is treated first as a control and later as a target of the intervention. The program to increase safety behavior conducted by Komaki, Barwick, and Scott illustrates this approach. After baseline data were collected on both departments, the program was introduced in the wrapping department eight and a half weeks before it was introduced into the makeup department. In this way, the makeup department served as a control for eight and a half weeks and then became a subject of the change program. The results showed that safety behavior increased in each department only after the intervention went into effect.
Behavior change programs can improve employee productivity. A key agent of change in all organizations at all levels is the immediate supervisor. With behavior management techniques, supervisors can increase employees' performance by teaching them how to work. Few schools teach people how to work; most people learn on the job. Goal-setting, self-reinforcement, self-monitoring, developing contingencies, and doing the least-liked work first are work behaviors. Learning by doing is a powerful and efficient educational process. Employees learn to come to work on time by coming to work on time and they learn to set goals by setting goals.
The challenge to the supervisor is to arrange the work environment so that employees are likely to do their work. With tardiness, the supervisor might add a contingent incentive; with goal-setting, the supervisor might guide the subordinate through the process many times. Learning to work is a continuous process: Accomplishing the objectives of one change program sets the stage for the next one. Once workers learn to be on time for an immediate continuous pay off, for example, they need to learn to continue to arrive on time for a less frequent payoff. The first intervention might involve receiving an incentive every time the employee is on time; the second might use intermittent incentives and goal-setting. Supervisors are in a key position-they can facilitate or inhibit the performance of those they supervise. And by learning to use the skills and techniques, they can become agents of change-and teachers.