Where reinforcement is the introduction of positive or removal of negative stimuli to increase a particular behavior, punishment is the introduction of negative or removal of positive stimuli to decrease a particular behavior.

Types of Punishment

There are four basic types of punishment.

Time out is the removal of the child from any positive stimuli. It is important that the motivation behind the behavior that one hopes to decrease is ascertained before a time out is considered (see the section of this site on Functional Behavior Analysis). For example, if a child screams each time he is asked to work, using a time out might actually encourage that behavior rather than reduce it. The child is likely screaming to escape from the request, so putting him in time out (removing him from the situation) is really giving him exactly what he wants. Consequently, the next time he wants to escape from an undesirable situation, the child might well try screaming again, as it was very successful the last time.

Extinction is the discouragement of a behavior that used to be reinforced by no longer reinforcing of it (essentially the ignoring of a behavior). Extinction is often successful at decreasing or even eliminating behaviors, but can be difficult. Typically, when one sets out on an extinction path for a particular behavior, the occurrence of that behavior actually increases for some period of time (how long depends on just how well-reinforced the behavior was). Consequently, if one is to maintain an extinction path for a behavior, it must be approached with a certain degree of vigilance, for reinforcement of the behavior along that initial path of escalation (or even after the peak) can cause the child increase his attempts to gain additional reinforcement for the behavior.

Response cost is a punishment based upon a token economy system. Tokens are taken away from the individual for occurrences of an undesirable behavior. For example, if I were to cheat on my taxes (a behavior the government finds undesirable) they would fine me a considerable sum (which I find rather undesirable). Consequently, I would, presumably, be less likely to repeat that behavior in the future. One must use caution when using response cost systems, however. In an educational setting, token economy systems can be very valuable tools, but they can be difficult to teach and to get the child to trust. It has been my experience the response cost programs can take a big toll on the effectiveness of a token economy system, they can confuse and greatly upset the child and can lead him to no longer trust the system. Consequently, one must weigh the relative value of possibly decreasing a particular behavior with the cost of possibly losing token economy systems as a teaching tool altogether.

Aversive stimuli are strongly negative stimuli introduced after an undesirable behavior. Spanking and scolding are the classic examples of aversive stimuli, but they need not be that severe. As a general rule, aversive stimuli are to be avoided, especially in school settings (unless with complete parental involvement under very well defined guidelines), as they often do more harm than good. Aversives as punishment often suppress even good behaviors for fear of punishment, the tend to encourage avoidance, escape, and strong emotional backlash. Aversives may also model inappropriate behaviors as solutions to problems, and may foster negativity between adult and child.

Alternatives to Punishment

As the purpose of discrete trial training is to teach positive behaviors, there should be no need for punishment when involved in a DTT session. The fact that problem behaviors are not being reinforced should be enough to discourage them in favor of the behaviors to which focus is being given. However, it may be necessary to lessen particular behaviors outside of the discrete trial session. The first step in this process is to ascertain the motivation behind the current behaviors (see the section of this site on Functional Behavior Analysis for more information). Once the function of the behavior has been determined, a plan can be created to teach better means of meeting those needs.

There are four basic alternatives to punishment for reducing difficult behaviors and encouraging more appropriate behaviors:

Differential Reinforcement of Other Behaviors (DRO) is the reinforcement of any behavior other than the difficult behavior.

Differential Reinforcement of Alternative Behaviors (DRA) is the reinforcement of behaviors which serve as alternatives to the difficult behavior, especially alternative means of communication. As an example, a child could be taught to present his teacher with a PECS symbol for a break from work instead of tantrumming to escape.

Differential Reinforcement of Incompatible Behaviors (DRI) is the reinforcement of behaviors which are incompatible with difficult behaviors, that is they are behaviors which the child can not be doing simultaneously. For example, if a child is constantly touching his neighbors during circle time, he could be reinforced for keeping his hands in his lap, or sitting on them.

Differential Reinforcement of Lesser Rates of Behavior (DRL) is the reinforcing of periods of time in which the child exhibits the behavior at a predetermined lesser rate. For example, if a child currently talks out approximately ten times every five minutes he is in the library with his class, he could be reinforced if he only talks out five times during that same five minutes.

In the next section we'll look at the Inter-trial Interval.

[ Previous | Contents | Next ]