MISAPPLICATIONS
339
Chapter Seventeen: Using Punishment: Time-out and Response Cost
How does time-out work to decrease a problem behavior?What are the two types of time-out What is response cost? How do you use it to decrease a problem behavior? Why is it important to use reinforcement procedures together with time-out or response cost
What issues must you consider when using time-out or response cost?
As discussed in Chapter 6, punishment is a basic behavioral principle. Punishment occurs when a behavior is followed by a consequence that results in a decrease in the future probability of the behavior. The consequence following the behavior may involve the presentation of an aversive stimulus or event (positive punishment) or the removal of a reinforcing stimulus or event (negative punishment). In both forms of punishment, the behavior is weakened.
A variety of punishment procedures can be used to decrease a problem behavior. However, punishment procedures typically are used only after functional interventions—extinction, differential reinforcement, and antecedent manipulations—have been implemented or considered. When these procedures are implemented and result in a decrease in the problem behavior, punishment procedures are unnecessary. However, if functional procedures are ineffective (or not completely effective) or if their use is limited or impossible for whatever reason, punishment procedures may be considered.
The use of punishment procedures can be controversial. Some people believe that using punishment, the contingent presentation of an aversive event or the removal of a reinforcing event, may violate the rights of the person being treated (e.g., see LaVigna&Donnelan, 1986). In addition, positive punishment involves presenting an aversive stimulus, which is often perceived to be painful or unpleasant; therefore, some people believe that punishment produces unnecessary pain or discomfort for the person receiving treatment. (Note, however, that an aversive stimulus is not defined by painful or unpleasant feelings. Rather, behavior modification adopts a functional definition, in terms of its effect on behavior: An aversive stimulus is any stimulus whose contingent presentation decreases the future probability of a behavior or whose contingent removal increases the future probability of a behavior; see, for example, Reynolds, 1968.)
For these and other reasons (see Chapters 6 and 18), punishment procedures usually are not the first choice of interventions for decreasing problem behaviors. If a punishment procedure is used, it is often a negative punishment procedure involving the removal of reinforcing events after a problem behavior. This chapter describes two common negative punishment procedures: time-out and response cost.
Time-out
Cheryl and the other kindergarten children were sitting around the table making figures out of clay, finger painting, and cutting shapes out of construction paper. After a little while, Cheryl threw one of her clay figures and smashed some figures made by other children. Seeing this, the teacher calmly walked up to Cheryl and said, Cheryl, come with me. She took Cheryl by the arm, and they walked to a chair across the room. When they got to the chair, she said, Cheryl, you can”t play when you throw things or break things. Sit here until I say you can play again. The teacher then walked back to the table and praised the other children for the figures they had made. After 2 minutes, the teacher walked back over to Cheryl and said, Cheryl, you can come back to the table and play now (Figure 17-1). When Cheryl came back and played without any further problems, the teacher talked to her and praised her for playing nicely. This procedure, in which Cheryl was removed from the reinforcing activity in the classroom for a few minutes contingent on an instance of the problem behavior, is called time-out. Once the teacher started using time-out, the rate of Cheryl”s problem behaviors decreased greatly.
FIGURE 17-1
When Cheryl engages in a problem behavior in the classroom, she has to sit off to the side and watch her classmates have fun for a few minutes. This procedure, a form of time-out called contingent observation, removes Cheryl from the reinforcement in the classroom for a few minutes contingent on the problem behavior.
342343
For about a year, 5-year-old Kenny had talked back to his parents and refused to do what they asked. Usually, Kenny was watching TV or playing a game when he engaged in these problem behaviors. Although his parents argued with him and gave him warnings about what might happen, he usually continued to watch TV or play and did not complete the task that was requested. His parents discussed the problem with a psychologist and decided to implement the following plan. First, when a parent wanted Kenny to do something, the parent went up to him, looked him in the eyes, and stated clearly what he or she wanted him to do. Second, the parent continued to stand there, and if Kenny did not do what was requested within a short time (10-15 seconds), the parent said, If you don”t do what I tell you, you have to sit in your room. The parent then took Kenny”s hand and walked him to his room. There were no toys, TV, or other recreational materials in Kenny”s room. The parent told him to stay there until told that he could leave. If Kenny argued, complained, or talked back during this process, the parent did not say anything to him in return. After a few minutes, the parent went to Kenny”s room and made the same request that Kenny had earlier refused. If Kenny complied with the request this time, the parent thanked him for performing the task and let him watch TV or resume playing. If he refused again, however, the parent told him that he would have to stay in his room longer and left him there. After a few more minutes, the parent returned and repeated the process until Kenny finally complied with the request. Finally, on occasions when Kenny complied with his parents” request without protest, the parents smiled and praised him enthusiastically.
These two examples illustrate the use of time-out (and other procedures) to decrease the occurrence of different problem behaviors. In each example, after the problem behavior, the child was removed from the reinforcing situation for a brief period. Playing with the clay and finger paints and interacting with the other children are reinforcing activities for Cheryl, and time-out involved removing her from the situation where these reinforcing activities were present. Watching TV or playing a game is reinforcing for Kenny, and time-out involved removing him from the opportunity to continue these activities.
What other behavioral procedures were used in conjunction with time-out in these examples?
Differential reinforcement of alternative behavior was used in both examples. When Cheryl was playing appropriately, the teacher reinforced this behavior with attention. When Kenny complied with his parents” requests, this behavior was reinforced with praise and the opportunity to continue the reinforcing activities that he was engaging in before the request. In addition, Kenny”s parents used a stimulus control procedure by standing directly in front of Kenny, looking him in the eyes, and stating the request clearly. The specific request, close proximity, and eye contact became a discriminative stimulus in the presence of which Kenny”s compliance was reinforced and his refusal behavior was punished (with time-out). Thus, whenever the parents make a request in this way, Kenny is less likely to refuse because refusal has been punished by time-out.
343
344
Types of Time-out
Time-out is defined as the loss of access to positive reinforcers for a brief period contingent on the problem behavior (Cooper, Heron, &Heward, 1987). The result is a decrease in the future probability of the problem behavior. Time-out as used in this section is short for time-out from positive reinforcement. There are two types of time-out: exclusionary and nonexclusionary.
The example of Cheryl illustrates nonexclusionary time-out. Cheryl remained in the classroom after the problem behavior but had to sit across the room from where the other children played, and was thus removed from the reinforcing activity. The case of Kenny illustrates exclusionary time-out. Contingent on the problem behavior, Kenny was taken out of the room where he was watching TV or playing. He was taken to a room where these reinforcers were not available.
Nonexclusionary time-out is most likely to be used when the person can be removed from the reinforcing activities or interactions while still remaining in the room, and the presence of the person in the room will not be disruptive to others in the environment. If either of these criteria cannot be met, exclusionary timeout would be used instead. For example, if Cheryl sat across the room in the time-out chair and disrupted the other students by continuing to engage in the problem behavior, nonexclusionary time-out would not be appropriate. Alternatively, if watching other children play were just as reinforcing for Cheryl as playing herself, nonexclusionary time-out would not be effective. For the procedure to be effective, the person must be removed from access to positive reinforcers. For Cheryl, exclusionary time-out could be implemented by having her sit in the principal”s office or in another room adjacent to the classroom for a few minutes each time she engaged in the problem behavior. In addition, nonexclusionary time-out might be effective if Cheryl was made to sit in a chair facing the wall.
Exclusionary Time-out
The person is removed from the room (the reinforcing environment) where the problem behavior occurred and is taken to another room. This removes the person from all sources of positive reinforcement.
Nonexclusionary Time-out
The person remains in the room while being removed from access to positive reinforcers
Using Reinforcement with Time-out
Whenever you use time-out (or any other punishment procedure), you should also use a differential reinforcement procedure. The time-out procedure decreases the rate of the problem behavior, and a differential reinforcement procedure increases an alternative behavior to replace the problem (differential reinforcement of alternative behavior [DRA]) or provides the reinforcer for the absence of the problem behavior (differential reinforcement of other behavior [DRO]), while at the same time applying extinction for the problem behavior. Because the time-out procedure 344345eliminates access to positive reinforcers contingent on the problem behavior, it is important for the person to have access to positive reinforcers through a DRA or DRO procedure. If you used time-out without a differential reinforcement procedure, there could be a net loss in reinforcement and the problem behavior could be more likely to reemerge after treatment.
Considerations in Using Time-out
To use time-out effectively, you must address a number of considerations.
What Is the Function of the Problem Behavior?
Time-out is appropriate to use with problem behaviors that are maintained by positive reinforcement involving social or tangible reinforcers. Time-out removes access to these and other positive reinforcers contingent on the problem behavior; as a result, the problem behavior is less likely to occur. In addition, the time-in environment (the environment where the problem behavior occurs) must consist of positively reinforcing activities or interactions for timeout to be effective. Removing the person from this environment is time-out from positive reinforcement only if the time-in environment is positively reinforcing and the time-out environment is not reinforcing or is less reinforcing (Solnick, Rincover, & Peterson, 1977).
Time-out is not appropriate to use with problem behaviors maintained by negative reinforcement or sensory stimulation (automatic reinforcement). Because timeout removes the person from the ongoing activities or interactions in the room, timeout would negatively reinforce any behavior that was maintained by escape (Plummer, Baer, & LeBlanc, 1977; Taylor & Miller, 1997). For example, suppose that a student engages in aggressive behavior in the classroom and this behavior is negatively reinforced by escape from the educational demands. If the teacher used time-out, removing the student from the classroom would negatively reinforce the aggressive behavior. Time-out negatively reinforces the problem behavior when the time-out environment is less aversive than the ongoing activities.
Likewise, when a problem behavior is maintained by sensory stimulation, time-out is not appropriate because it would not function as time-out from positive reinforcement. The person would be removed from the activities or interactions in the time-in environment and would have the opportunity to engage in the problem behavior while alone in the time-out area (Solnick et al., 1977). Because the problem behavior is reinforced automatically by the sensory stimulation it produces, time-out would be reinforcing: The person would have the opportunity to engage in the automatically reinforced behavior without interruption
FOR FURTHER READING: Functional Considerations in Using Time-out
Studies by Plummer and colleagues (1977), Solnick and colleagues (1977), and Taylor and Miller (1997) established that the functional context in which time-out is used influences the effectiveness of time-out. Because time-out involves removal of the individual from a reinforcing environment for a brief period contingent on the problem behavior, time-out will not work if the time-in environment is aversive, devoid of reinforcers, or less reinforcing than the time-out environment. For example, Plummer and colleagues (1977) evaluated the effects of time-out for disruptive behavior that occurred during classroom instruction. 345346They showed that disruptive behavior increased when time-out was used because time-out provided escape from instruction (negative reinforcement). Solnick and colleagues likewise showed that time-out was effective when the time-in environment was enriched (filled with reinforcers) but was not effective when the time-in environment was impoverished (devoid of reinforcers). Solnick and colleagues also showed that time-out did not work when an individual had the opportunity to engage in self-stimulatory behavior during time-out; thus, the time-out environment was more reinforcing than the time-in environment. Finally, Taylor and Miller (1997) showed that time-out was effective for problem behaviors maintained by attention from teachers but was not effective for problem behaviors maintained by escape from academic tasks. When problem behaviors were maintained by attention, time-out removed attention and thus functioned as negative punishment (thereby decreasing the problem behavior). When problem behaviors were maintained by escape, time-out provided escape and thus functioned as negative reinforcement (thereby increasing the problem behavior).
Is Time-out Practical in the Given Situation?
Time-out is practical when the change agents can implement the procedure successfully and the physical environment is conducive to its use. In the time-out procedure, the person often is removed from the room or from the area of the room where the problem behavior occurs. The change agent implementing time-out often must physically escort the client to the time-out room or area. In some cases, the client may resist when being escorted to timeout areas. If the resistance involves physical confrontation or aggression, especially if the client is a large person (e.g., an adult with intellectual disability or a psychiatric disorder), the change agent may not be able to implement the procedure. This factor must be considered before time-out is chosen as a treatment.
The second practical consideration is whether there is an appropriate room or area to use for time-out. For exclusionary time-out, another room or a hallway can be used. However, the time-out area must be a place where the client does not have access to any positive reinforcers. If a child is sent to his or her room for time-out and the room has a TV, video games, and toys, the room is not an appropriate place for time-out. If other people interact with the client during time-out, the time-out area is not appropriate. For example, if a student is sent to sit in a hallway where his or her friends are hanging out, time-out will not be effective. If no room or area exists where the client can be removed from positive reinforcers, time-out cannot be implemented.
Sometimes a room is built or an existing room is modified specifically for use as a time-out room. Such a room should be safe (free of sharp or breakable objects), well-lighted (with a ceiling light that cannot be broken), and barren (empty except for a chair). In addition, there should be an observation window so that the client can be observed during time-out. A one-way observation window is best so that the client cannot see the observer. Finally, the room should not have a lock so that the client cannot lock out the change agent and the change agent cannot lock in the client. This precaution safeguards against the misuse of a time-out room. It would be a misuse of time-out for the change agent to lock the door and leave a client unattended in a time-out room.
Is Time-out Safe?
As noted earlier, the time-out room must not contain any objects that clients could use to hurt themselves. In addition, although the change agent must not interact with clients during time-out, the change agent should 346347observe them throughout the duration of time-out to ensure that they do not harm themselves. This is especially important for clients who engage in violent, aggressive, or self-injurious behavior.
Is the Time-out Period Brief?
Time-out is a brief loss of access to positive reinforcers. The problem behavior should result in an immediate removal from the reinforcing time-in environment. However, the client should be returned to the time-in environment as soon as possible and allowed to resume normal activities (whether educational, vocational, or recreational). Time-out duration is typically 1-10 minutes. However, if the client is engaging in problem behaviors in the time-out area at the end of the timeout period, time-out is extended for a brief time (typically 10 seconds to 1 minute) until the client is no longer engaging in problem behaviors. The absence of the problem behavior is required at the end of time-out so that the termination of time-out does not negatively reinforce the problem behavior. This extension of time-out is called a contingent delay. Although an earlier study by Mace, Page, Ivancic, and O”Brien (1986) found that time-out was equally effective with or without the contingent delay, a more recent study by Erford (1999) showed that time-out with a contingent delay was more effective than time-out without a contingent delay. Although research findings are mixed, a contingent delay is recommended so that problem behavior is not inadvertently reinforced by escape from time-out. When the client is released from time-out, the change agent should identify the desirable behavior that will be reinforced in the time-in environment.
Can Escape from Time-out Be Prevented?
Whether using exclusionary or nonexclusionary time-out, the change agents should prevent the client from leaving the timeout room or area before the end of the time-out interval. If implemented correctly, time-out is aversive for the client, so the client may try to leave. However, for time-out to be effective, the client must not leave until the interval is up. For example, if a parent is using a time-out chair with a 5-year-old, the parent must keep the child in the chair during time-out. If the child gets up, the parent (who is standing near the child in the chair) should calmly give the child the instruction to sit back down. If the child does not comply or if the child gets up repeatedly, the parent should use physical guidance to keep the child in the chair. This may vary from a hand on the shoulder to physically restraining the child in the chair (McNeil, Clemens-Mowrer, Gurwitch, &Funderburk, 1994). When a time-out room is being used, the parent must return the child to the room if the child leaves prematurely; alternatively, the parent may hold the door closed as the child tries to open it. In either case, it is important to avoid a struggle, which may be reinforcing for the child, and thus make time-out less effective. If the parent cannot prevent escape from time-out or cannot avoid a reinforcing struggle, time-out should not be used.
Can Interactions Be Avoided during Time-out?
Time-out must be implemented calmly and without any emotional response from the change agent. In addition, while taking the client to time-out or during time-out, the change agent must not interact socially with the client. Reprimands, explanations, or any other form of attention must be avoided during time-out because they lessen its effectiveness. 347348For example, if a child sitting in a time-out chair whines, cries, calls the parent names, or says, I hate you, or if the child pleads to get out of the chair or promises to be good, the parent should stand nearby and ignore this behavior until the time-out interval is up. If the child resists going to the time-out chair or room, the parent must not scold the child or try to talk the child into complying. The parent should simply provide the degree of physical guidance that is necessary to get the child to the time-out room or chair.
Is Time-out Acceptable in the Given Situation?
In some treatment settings, such as programs for people with intellectual disabilities, rules and regulations govern the use of timeout and other punishment procedures. Before deciding whether to use time-out, you must be certain that the procedure is acceptable in the particular treatment environment. In addition, when working with parents, it is important to assess the degree to which they find time-out to be an acceptable procedure. Although acceptability may be increased through rationales and explanations about a particular treatment, ultimately the parents must accept the use of time-out if they are to implement it with their child.
Considerations in Using Time-out
What is the function of the problem behavior?
Is time-out practical in the circumstances?
Is time-out safe?
Is the time-out period brief?
Can escape from time-out be prevented?
Can interactions be avoided during time-out?
Is time-out acceptable in the circumstances?
Research Evaluating Time-out Procedures
Numerous studies have demonstrated the effectiveness of time-out with children and with people with intellectual disabilities (Adams & Kelley, 1992; Bostow& Bailey, 1969; Handen, Parrish, McClung, Kerwin, & Evans, 1992; Hobbs, Forehand, & Murray, 1978; Mace et al., 1986; McGimsey, Greene, &Lutzger, 1995; Roberts & Powers, 1990; Rolider& Van Houten, 1985; Taylor & Miller, 1997).
Porterfield, Herbert-Jackson, and Risley (1976) and Foxx and Shapiro (1978) investigated two variations of nonexclusionary time-out. Porterfield and her colleagues evaluated time-out with young children who engaged in aggressive and disruptive behaviors in a day care program. When a child engaged in a problem behavior, the caregiver took the child outside the play area and had the child sit down on the floor and watch the other children play. After the child sat there for about a minute with no toys, activities, or interactions, the caregiver allowed the child to return to the play area. The caregiver also praised the other children for playing appropriately. Porterfield called this procedure contingent observation because, contingent on the occurrence of the problem behavior, the child had to sit and watch the other children play appropriately. The procedure decreased the level of disruptive and aggressive behavior of the children in the day care program.
Foxx and Shapiro (1978) worked with five boys with intellectual disabilities who engaged in a variety of problem behaviors (hitting, throwing objects, yelling, getting out of seat, banging objects) in a special education classroom. The boys sat around a table where the teacher worked with them on various educational activities. The teacher delivered edible and social reinforcers to each student at intervals of about 2 minutes when the student was not exhibiting a problem behavior. During the time-in condition, each student wore a different-colored ribbon around his neck. When the student engaged in a problem behavior, the teacher removed the ribbon and put it around her neck as a signal that time-out was in effect for that student. As long as the student was not wearing his ribbon, he could not engage in any activities and could not receive the reinforcers. The time-out lasted for 3 minutes. Use of this nonexclusionary time-out procedure resulted in decreases in problem behaviors for all five boys.
Mathews, Friman, Barone, Ross, and Christophersen (1987) worked with mothers and their 1-year-old children. The researchers instructed the mothers to use exclusionary time-out when their children engaged in dangerous behavio