Noncontingent Reward (Reinforcement)

Description of the Strategy

Noncontingent reward (NCR)* is a procedure that, as its name suggests, involves delivering rewards independent of the occurrence of any specified behavior. Sometimes NCR is used to make a particular setting more attractive. That is, offering a high rate of rewards reinforces entering and staying in that setting. However, once in the setting, receipt of the rewards does not depend on engaging in any particular response and, in fact, may alter previously established patterns of behavior. Indeed, the most common and well-studied use of NCR is as a procedure for decreasing problem behavior. To illustrate, imagine a child diagnosed with autism who engages in self-injurious behavior (e.g., head banging), maintained by social reinforcement (i.e., the child receives attention following head banging). A typical NCR procedure to decrease [Page 920]head banging would involve the regular provision of attention regardless of head banging.

Research Basis

The general description of the NCR procedure leaves unspecified a number of parameters that appear important to consider in employing NCR. The first is to decide the type of schedule according to which the rewards will be offered. Will they be continuously available, delivered at regular time intervals (a fixed-time schedule), or at irregular time intervals (a variable-time schedule)? Building on the above example, attention might be provided continuously, every 60 seconds, or at varying intervals averaging 60 seconds. Fixed-time schedules have been the most widely used to date, but all three varieties have been used effectively, with continuous reward often employed when the problem behavior is maintained by automatic/ sensory reinforcement.

In addition to determining how rewards will be scheduled, the frequency or magnitude of reward delivery must be determined. Some, but not all, of the available research has shown that higher rates or magnitudes have led to greater reductions in problem behavior. Thus, the generally preferred strategy appears to be to begin the intervention using a relatively high rate/magnitude, which is gradually reduced after the problem behavior has declined.

There are a number of additional variations on the basic procedure that have been explored in the research literature and which a practitioner needs to consider. When possible, it appears generally preferable to use the reinforcer maintaining the problematic behavior (determined through a functional analysis) as the noncontingent reward. In the previous example, attention was maintaining head banging, and thus attention was chosen as the noncontingent reward. In addition, a decision has to be made whether to use NCR with or without extinction. For example, NCR could be introduced while continuing to provide attention following head banging, or head banging could be simultaneously put on extinction. Most interventions employ NCR and extinction; however, both NCR with and without extinction have demonstrated some efficacy, and studies directly comparing the two procedures have yet to reveal clear superiority of one over the other.

It is not always possible to identify the reinforcer(s) maintaining the problematic behavior, and even when identified, it is not always feasible to provide rewards that approximate the functional reinforcer (e.g., in the case of some automatic/sensory reinforcers). In such cases, one might have to use an arbitrary reward (e.g., food or toys) determined, when possible, on the basis of a preference assessment. Whether noncontingent delivery of arbitrary rewards is as useful in reducing problem behavior as using the reinforcer maintaining the behavior remains to be fully determined. That said, arbitrary rewards have been successfully used in several studies and appear most effective when a wide variety of high preference items are utilized.

...