discussion question

A review of “noncontingent” reinforcement as treatment for the aberrant behavior of

individuals with developmental disabilities

James E. Carr*,1, Sean Coriaty, David A. Wilder2, Brian T. Gaunt3, Claudia L. Dozier4, Lisa N. Britton5,

Claudia Avina, Curt L. Reed University of Nevada, Reno, NV 89557-0062, USA

Abstract

The term noncontingent reinforcement (NCR) refers to the delivery of an aberrant behavior’s known reinforcer on a response-independent basis. The typical result is a decrease in responding from baseline (i.e., reinforcement) levels. NCR has become one of the most reported function-based treatments for aberrant behavior in the recent literature. The purpose of this review is to briefly discuss the history of the procedure and summarize the findings from the treatment research literature. The review is organized into the following sections: (a) basic research on NCR, (b) NCR as a control procedure, (c) NCR as a function-based treatment, (d) considerations in the programming of NCR schedules, (e) behavior-change mechanisms underlying NCR effects, and (f) directions for future research. © 2000 Elsevier Science Ltd. All rights reserved.

1 James Carr is now at Department of Psychology, Western Michigan University, 1903 W. Michigan Ave., Kalamazoo, MI 49008-5439.

2 David Wilder is now at The University of the Pacific. 3 Brian Gaunt is now at The University of South Florida. 4 Claudia Dozier is now at The University of Florida. 5 Lisa Britton is now at Spectrum Center. * Corresponding author. Tel.:11-616-387-4925; fax:11-616-387-4550. E-mail address:jim.carr@wmich.edu (J.E. Carr).

Research in Developmental Disabilities 21 (2000) 377–391

0891-4222/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 8 9 1 - 4 2 2 2 ( 0 0 ) 0 0 0 5 0 - 0

1. Introduction

One of the most significant advances in the treatment of aberrant behavior of individuals with developmental disabilities has been the development of system- atic methods for determining behavioral function. The advent of functional assessment methods has enabled researchers and practitioners to design inter- ventions that are based on a behavior’s maintaining variable(s), rather than its topography. Function-based approaches are considered preferable to their non- function-based counterparts for several reasons. First, function-based interven- tions necessarily emphasize reinforcement-based, as opposed to punishment- based, procedures. In addition, function-based interventions are directed at the response-reinforcer relationship, thus weakening a problematic contingency as opposed to simply overpowering it. Finally, function-based interventions often involve the establishment of an adaptive response-reinforcer relationship, which may result in increased maintenance of treatment gains.

Iwata, Vollmer, Zarcone, and Rodgers (1993) described three general classes of function-based interventions. The first class, the modification of establishing opera- tions (EOs), includes antecedent manipulations designed to either weaken the rein- forcer for the aberrant behavior or strengthen that of an alternative behavior (for a review, see Wilder & Carr, 1998). The second class, extinction, involves withholding the reinforcer that maintains the aberrant behavior. Finally, in behavioral replacement procedures, the aberrant behavior’s reinforcer is provided contingent upon an alter- native behaviorand withheld for the aberrant behavior. In recent years, one of the most widely researched group of treatments for aberrant behavior has been noncon- tingent reinforcement (NCR). NCR procedures are often conceptualized as EO manipulations, although extinction sometimes plays a role in their effects. The basic premise of NCR is that an aberrant behavior’s reinforcer is delivered to the individual on a response-independent basis. Through the operation of several behavioral pro- cesses, a subsequent behavior reduction is usually observed.

Before the more substantive areas of this review are presented, a brief discussion of the termnoncontingent reinforcementis warranted. The quotes in the title of the article are a hint that the term is problematic. As has previously been discussed in the literature, NCR is a misnomer for several reasons. First, by definition, reinforcement is acontingentprocess. Second, no behavior is strength- ened as a result of NCR, compared to reinforcement procedures. Third, NCR is an imprecise description as it refers to many procedurally distinct treatments. With all of these problems with the term, one would surmise that researchers would have adopted a more precise vocabulary with which to refer to these procedures. However, this has not yet happened, and with the increasing fre- quency at which the term is being reported in the literature, we believe that a change is unlikely to occur. Therefore, throughout this review, the term NCR will be used much as it is used in the current treatment literature. We refer the reader to several published commentaries on this issue for more in-depth discussion, as well as for potential alternatives to current terminology (Carr, 1996; Poling & Normand, 1999; Vollmer, 1999).

378 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

The purpose of this review is to briefly discuss the history of NCR and summarize the recent findings from the research literature. First, the basic research on NCR is briefly summarized, followed by a short discussion of NCR’s role as a methodological control procedure. Next, NCR is discussed as a func- tion-based treatment for aberrant behavior. We then discuss the many variables that should be considered when programming NCR schedules. Next, we discuss the behavior-change mechanisms responsible for NCR effects. Finally, sugges- tions for future research are presented.

1.1. Basic research on NCR

Although Skinner (1948) first addressed the possibilities of fixed-time (FT) schedules of reinforcement in his seminal article “Superstition in the Pigeon,” it was not until the 1960s that researchers began to systematically study NCR procedures. Since then, there have been dozens of articles published in the basic literature whose primary purposes have been to (a) examine the procedural and functional properties of NCR, and (b) use NCR as a control procedure for studying other behavioral phenomena (e.g., behavioral momentum). Since the purpose of the current paper is to review NCR procedures as they relate to clinical application, a comprehensive review of the basic NCR literature is not possible. Below, however, is a brief summary of some of the early findings from the basic literature that were important to the development of our current technology.

Zeiler (1968) evaluated variable-time (VT) and FT schedules in pigeons that had previously acquired an operant response (i.e., key pecking). The results showed that FT schedules produced consistent decreases in responding, while VT schedules produced ones that were more erratic. Rescorla and Skucy (1969) compared extinction with the response-independent delivery (i.e., VT schedules) of food in rats. The authors found that extinction (by omission) decreased responding more effectively than response-independent food delivery, although both procedures resulted in significant decreases in response rates. Lachter, Cole, and Schoenfeld (1971) evaluated both dense and lean FT schedules in pigeons and found that, although both were effective reductive procedures, the dense schedules produced the greater reductions. The aforementioned studies and others (e.g., Alleman & Zeiler, 1974; Calef et al., 1989; Dickinson & Charnock, 1985; Edwards, Peek, & Wolfe, 1970; Halliday & Boakes, 1971; Job, 1988; Oakes, Rosenblum, & Fox, 1982) illuminated a consistent phenomenon: a change from response-dependent to response-independent reinforcement consistently produces a response reduction. This finding led to NCR being adopted as an experimental control procedure as an alternative to extinction.

1.2. NCR as a control procedure

Based on findings from the basic literature, researchers introduced NCR in the applied literature as an experimental control procedure. That is, reinforcers were

379J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

delivered independent of the target response in either FT or VT schedules to evaluate the effects of reinforcement-based procedures. Previously, extinction had been employed as the modal control procedure; however, extinction was shown to produce a possible confound. When extinction procedures are used, not only is the response-reinforcer relationship terminated, but thepresentationof the stimulus is also removed from the individual. This type of control (i.e., extinction) does not allow for the delineation of the effects of stimulus presen- tation and the response-reinforcer relationship (Lachter, 1980). In contrast, NCR procedures interrupt the response-reinforcer relationship while still presenting the stimulus to the individual, which allows for the examination of the response- reinforcer relationship independent of stimulus-presentation effects.

Numerous studies in the applied literature have since employed NCR as a control for reinforcement procedures (e.g., O’Neill & Morris, 1979; Sheppard, 1969). Below are brief descriptions of two representative studies. Hart, Reynolds, Baer, Brawley, and Harris (1968) reported one of the first demonstrations of the use of NCR as a control procedure with humans. The researchers used a reversal design to evaluate the effects of contingent adult social reinforcement on the cooperative play behavior of a 5-year-old girl. In addition to a contingent attention condition, the authors employed a noncontingent attention control procedure. Increased cooperative play was observed in the contingent attention condition, but not in the noncontingent attention condition. These results sug- gested that the contingency between the response and the reinforcer was respon- sible for the effects of the intervention. Similarly, Horner (1980) used an NCR control procedure to assess the effects of differential reinforcement, in conjunc- tion with environmental enrichment, on the adaptive and inappropriate behavior of five individuals diagnosed with profound mental retardation. The NCR pro- cedure consisted of maintaining the environmental enrichment procedure while providing social reinforcement independent of the individuals’ adaptive and inappropriate behavior. The results suggested that the differential reinforcement procedure (i.e., the contingency between the response and the reinforcer) was responsible for the decrease in inappropriate behavior and the increase in adap- tive behavior.

More recently, experimental functional analysis methods (Iwata, Dorsey, Slifer, Bauman, & Richman, 1994) have incorporated NCR as a control proce- dure. Functional analyses typically involve observing the occurrence of aberrant behavior during several experimental test conditions. The most commonly re- ported conditions are (a) contingent attention, (b) escape from instructional demands, (c) contingent materials (i.e., tangible items), (d) alone or no interac- tion, and (e) unstructured play (the NCR control condition). During the “control” or “play” condition, the individual receives frequent attention that is usually delivered on an FT schedule (e.g., FT 30s). In addition, the individual has access to preferred stimuli and/or leisure activities and is not presented with instruc- tional demands. This condition serves as a control procedure for the (a) presence of the experimenter, (b) availability of preferred stimuli and/or leisure activities, (c) absence of the presentation of demands, and (d) delivery of attention.

380 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

Responding in the control condition is then compared to responding in the other experimental conditions to assess behavioral function. Iwata et al. (1994) pub- lished an epidemiologic analysis of 152 functional analyses of the self-injurious behavior (SIB) of individuals diagnosed with developmental disabilities. The authors reported that NCR (via the control or play condition) was successful in reducing SIB in 127 out of the 152 cases (i.e., 83.6%).

Although NCR has proved useful as an experimental control procedure, there are times when other control procedures may be preferable. For example, with high-rate behaviors it is possible that reinforcers delivered noncontingently will be temporally contiguous with the occurrence of the target behavior. This may result in adventitious reinforcement and, therefore, increase the occurrence of the target behavior. On such occasions, the experimenter might consider using another control procedure, such as differential reinforcement of zero rates of behavior (DRO; Cooper, Heron, & Heward, 1987).

1.3. NCR as a function-based treatment

Boe (1977) and Thelen (1979) were the first to report the use of NCR as treatment for aberrant behavior. Boe used the VT delivery of food to reduce the aggression of a group of women diagnosed with mental retardation. Thelen used informally delivered noncontingent attention to reduce the aggressive tantrums of an 8-year-old girl. Although these were the first reports of NCR as treatment for aberrant behavior, the efforts cannot necessarily be considered “function-based,” as behavioral function was unknown in each report. More recently, NCR has been defined as “a response-independent or time-based delivery of stimuli with known reinforcing properties” (Vollmer, Iwata, Zarcone, Smith, & Mazaleski, 1993, p. 10). The phraseknown reinforcing propertiesis especially important. Given the myriad behaviors in which an individual can engage, NCR cannot be implemented in afunction-basedmanner unless the reinforcing stimulus is identified for the aberrant behavior.

The recent literature has identified numerous benefits of implementing NCR as a treatment for aberrant behavior. First, as previously stated, NCR is usually a function-based procedure that addresses the response-reinforcer relationship. Second, NCR has been shown to produce greater, or at least comparable behavior reductions relative to DRO (Vollmer et al., 1993), differential reinforcement of alternative behavior (DRA; Kahng, Iwata, DeLeon, & Worsdell, 1997), and extinction (Vollmer et al., 1998). Third, NCR has been shown to result in a higher rate of reinforcer delivery compared to other comparable procedures (e.g., DRO; Britton, Carr, Kellum, Dozier, & Weil, in press; Vollmer et al., 1993). Fourth, FT schedules have been reported to be relatively easy to implement compared to other procedures. Although the delivery of stimuli at fixed times intuitively seems easier than implementing, for example, a DRO procedure, no social validity assessments have been reported to support this claim. Fifth, NCR has been shown to produce less extinction-induced behavior (e.g., aggression,

381J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

response bursts) compared to other treatments (Vollmer et al., 1993; Vollmer et al., 1998).

In addition to the above benefits associated with NCR, recent treatment studies have demonstrated the generality of NCR across behavioral topography and function. NCR has been used to effectively treat aberrant behavior main- tained by attention (e.g., Hagopian, Fisher, & Legacy, 1994; Mace & Lalli, 1991; Vollmer et al., 1993), access to materials (e.g., Lalli, Casey, & Kates, 1997; Marcus & Vollmer, 1996; Smith, Lerman, & Iwata, 1996), escape and/or avoid- ance of instructional demands (e.g., Kahng et al., 1997; Vollmer, Marcus, &

Table 1 Summary of noncontingent reinforcement treatment studies

Author(s) Year n Age Range Primary Diagnoses

Boe 1977 29 8–29 Profound-Severe MR Britton et al. in press 3 7–41 Profound-Severe MR, Autism Carr & Britton 1999 1 32 Moderate MR Coleman & Holmes 1998 3 4 Autism, PDD Derby et al. 1996 1 12 Profound MR Fischer et al. 1997 2 34–44 Profound MR Fisher et al. 1996 1 4 Severe-Moderate MR, Autism Fisher et al. 1999 3 9–13 Moderate-Mild MR Hagopian et al. 1994 4 5 Severe-Mild MR, PDD Hagopian et al. 2000 1 26 Moderate MR Hanley et al. 1997a 2 11–16 Severe-Moderate MR Hanley et al. 1997b 2 4–8 Mild MR, Cerebral Palsy Kahng et al. 1997 3 29–50 Profound-Severe MR Lalli et al. 1997 3 3–9 Severe-Mild MR Lalli et al. 1998 1 10 Severe MR Luiselli 1994 1 8 Posttraumatic Neurological

Impairment Mace & Lalli 1991 1 46 Moderate MR Mace et al. 1998 1 7 Moderate MR, Autism Marcus & Vollmer 1996 3 4–5 Profound-Moderate MR,

Autism Persel et al. 1997 1 40 Traumatic Brain Injury Piazza et al. 1996 1 17 Severe MR Piazza et al. 1997 1 8 Mild MR Piazza et al. 1998 3 4–17 Profound-Moderate MR,

Autism Roscoe et al. 1998 3 20–35 Profound-Moderate MR Sigafoos & Pennell 1995 1 10 Severe MR Smith et al. 1996 1 32 Profound MR Sprague et al. 1997 2 9–20 Severe MR Thelen 1979 1 8 none Vollmer et al. 1993 3 32–42 Profound-Severe MR Vollmer et al. 1995 2 4–18 Profound MR Vollmer et al. 1997 1 13 Severe MR Vollmer et al. 1998 3 6–22 Severe-Moderate MR Wilder et al. 1997 1 46 Profound MR

(continued on next page)

382 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

Ringdahl, 1995), and automatic reinforcement (e.g., Roscoe, Iwata, & Goh, 1998; Sprague, Holland, & Thomas, 1997). Additionally, recent studies have reported that NCR can be implemented with “arbitrary” stimuli identified from preference assessments (Fischer, Iwata, & Mazaleski, 1997; Hanley, Piazza, & Fisher, 1997). This finding is useful for cases in which behavioral function cannot be determined or the maintaining reinforcer cannot be withheld (e.g., some automatically reinforced behaviors).

Table 1 (continued)

Aberrant Behavior(s) Function(s) NCR Form Schedule Thinning

Aggression Unknown Variable-Time No Aggression, SIB Attention, Materials Fixed-Time Yes Inappropriate Speech Attention Fixed-Time Yes Aggression, Disruption Escape Fixed-Time Yes SIB, Self-Restraint Attention Continuous No SIB Attention, Materials Fixed-Time Yes Aggression, Disruption Attention Fixed-Time Yes Aggression, Disruption, SIB Attention Fixed-Time No Aggression, Disruption, SIB Attention Fixed-Time Yes Excessive Medical Complaints Attention Fixed-Time Yes Aggression, Disruption, SIB Attention Continuous No Aggression, Disruption Attention, Escape Fixed-Time No SIB Attention, Escape Fixed-Time No Aggression, SIB Materials Fixed-Time Yes SIB Attention Fixed-Time Yes Stereotypy Unknown Continuous No

Inappropriate Speech Attention Variable-Time No SIB Escape, Materials Fixed-Time No Aggression, SIB Materials Fixed-Time Yes

Aggression, SIB Attention Fixed-Time No Pica Automatic Continuous No Aggression, Disruption Attention, Escape Continuous No Pica Attention, Automatic Continuous No

SIB Automatic Continuous No SIB Automatic Continuous No SIB Materials Continuous No SIB, Stereotypy Automatic Variable-Time No Aggression Unknown Unknown No SIB Attention Fixed-Time Yes SIB Escape Fixed-Time Yes Aggression Materials Fixed-Time Yes Aggression, Disruption Attention, Escape Fixed-Time Yes Rumination Unknown Fixed-Time No

Note: MR 5 mental retardation, PDD5 pervasive development disorder, SIB5 self-injurious behavior. “Escape” refers to behaviors that are maintained by negative reinforcement contingencies. “Unknown” is reported in the function column for studies that did not report a systematic functional assessment method.

383J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

NCR has also been shown to be an effective treatment for a variety of aberrant behaviors including aggression (e.g., Vollmer, Ringdahl, Roane, & Marcus, 1997; disruption (e.g., Fisher, Ninness, Piazza, & Owen-DeSchryver, 1996), inappropriate speech (e.g., Carr & Britton, 1999), pica (e.g., Piazza et al., 1998), rumination (e.g., Wilder, Draper, Williams, & Higbee, 1997), SIB (e.g., Fischer et al., 1997), and stereotypy (e.g., Sprague et al., 1997). More detailed summaries of the aforementioned studies and others can be found in Table 1.

1.4. Considerations in the programming of NCR schedules

In this section of our review, we present brief discussions of several consid- erations that are important when programming NCR schedules. These consider- ations include (a) selecting NCR schedule parameters (i.e., form, density, rein- forcer magnitude, schedule-thinning method), (b) addressing the possibility of adventitious reinforcement, and (c) combining NCR with adjunct procedures.

1.4.1. Schedule form As previously stated, NCR refers to a wide range of procedurally distinct

treatments. For example, NCR includes FT, VT, random-time (RT), and contin- uous-presentation schedules. In addition, any schedule that delivers the reinforcer for a behavior on a response-independent basis can be considered an NCR procedure. The majority of the NCR treatment studies involve the implementa- tion of FT schedules. The results have proved successful, and as previously mentioned, FT schedules offer the potential benefit of ease of implementation. Throughout this review, we use NCR to refer to FT schedules, unless otherwise stated. Only one study so far has reported the use of VT schedules to reduce aberrant behavior (Mace & Lalli, 1991). To date, there have been no comparisons between FT and VT schedules with humans, nor has there been an evaluation of RT schedules. NCR procedures that involve the continuous presentation of a behavior’s reinforcer have been successfully reported, primarily with behaviors maintained by automatic reinforcement (e.g., Roscoe et al., 1998).

1.4.2. Schedule density and reinforcer magnitude Hagopian et al. (1994) demonstrated that denser NCR schedules produced greater

reductions than leaner schedules. This finding was a replication of a similar finding reported by Lachter et al. (1971) in the basic literature. Most NCR treatment studies report initial dense schedules that are thinned to leaner ones in order to capitalize on this effect. Several studies have shown that schedules denser than the baseline reinforcement rate can be effective in reducing behavior (e.g., Ringdahl & Vollmer, 1998; Wilder, Carr, & Gaunt, in press). In a study that incorporated both basic and applied features (i.e., a bridge study), Carr, Bailey, Ecott, Lucker, and Weil (1998) demonstrated that NCR schedules with high-magnitude reinforcers produced greater reductions than schedules with lower-magnitude reinforcers. The studies by Hago-

384 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

pian et al. (1994) and Carr et al. (1998) demonstrate the importance of both schedule density and reinforcer magnitude to the effectiveness of NCR procedures. These parameters should both be considered when developing an NCR treatment plan, especially with respect to behavioral function and frequency, and the practical resources required to implement the plan.

1.4.3. Schedule thinning

After the functional assessment and baseline periods, NCR is typically initially administered in a dense, often continuous, manner. Once the aberrant behavior is reduced to acceptable levels, the schedule is then thinned to a leaner version. If the behavior remains at acceptable levels, the schedule is thinned again. If the behavior increases above the level of acceptability, the schedule is usually returned to its previous version. This process is typically repeated until the NCR schedule is lean enough to be implemented in a practical manner. The modal terminal point for the NCR thinning process in the research literature has been 5 min (this is further discussed below in Directions for future research). Systematic methods for thinning NCR sched- ules have not yet been empirically established. In a related matter, however, Lalli et al. (1997) established an effective method for programming initial NCR schedules. The authors demonstrated that initial NCR schedules based on the mean latency to the first aberrant behavior during baseline were effective in reducing behavior. This method quickens the thinning process, as the resulting initial schedule is typically less-than-continuous. For a more detailed discussion of schedule thinning, we refer the reader to a review by Tucker, Sigafoos, and Bushell (1998).

1.4.4. Combining NCR with extinction

Another consideration in the programming of NCR schedules is whether to incorporate an extinction component. In most of the studies reported in the treatment literature, NCR has been implemented concurrently with extinction. In other words, not only was the aberrant behavior’s reinforcer delivered on a response-independent basis, but the behavior no longer produced access to that reinforcer. To date, there have been only a few studies that have successfully implemented NCR without extinction (e.g., Fischer et al., 1997; Fisher et al., 1999; Lalli et al., 1997). Hence, it is largely unknown how important extinction is to the success of NCR procedures. However, to the extent that NCR achieves its effects via an extinction process (seeBehavior-change mechanisms underly- ing NCR effectsbelow), it is probably reasonable to include extinction concur- rently with NCR.

1.4.5. Adventitious reinforcement

One of the potential side effects associated with NCR is the risk of adventi- tious, or accidental, reinforcement of the aberrant behavior. As Skinner demon- strated in 1948, it is possible for the FT delivery of a stimulus to reinforce

385J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

behavior unintended by the deliverer. One would surmise that aberrant behavior that occurs at a high frequency would be more susceptible to adventitious reinforcement than a lower-frequency behavior. Using the same logic, a denser NCR schedule would be more likely to produce adventitious reinforcement than a leaner schedule. However, a review of the NCR treatment literature reveals many instances of successful treatment with dense schedules and high-rate behaviors. To date, the variables that produce adventitious reinforcement have not yet been delineated. However, two procedures can be used to prevent adventitious reinforcement from occurring. Vollmer et al. (1997) used an omis- sion contingency (i.e., fixed-time, nonresetting, momentary DRO) to reduce aberrant behavior that had been adventitiously reinforced during NCR. In this procedure, reinforcer deliveries are omitted if the aberrant behavior has just occurred. Similarly, Britton et al., in press reported the use of NCR with a stimulus delay procedure to prevent adventitious reinforcement. With this vari- ation of NCR, if the aberrant behavior occurs within 10s of reinforcer delivery, the delivery is postponed in 10-s increments until the aberrant behavior has not occurred for 10s. Thus, the possibility of adventitious reinforcement is elimi- nated, while the amount of stimulation (i.e., reinforcement) received by the individual is retained.

1.4.6. Combining NCR with adjunct procedures One of the limitations associated with NCR is that, although an aberrant behavior

can be successfully treated, an appropriate alternative response is not strengthened. Marcus and Vollmer (1996) and Goh, Iwata, and DeLeon (2000) reported studies that addressed this limitation. Both studies evaluated the effects of NCR combined with a DRA procedure in terms of behavioral reduction and alternative response acquisition. The NCR/DRA treatment packages were successful in reducing aberrant behavior, and once the initial continuous schedule was thinned, individuals success- fully acquired the alternative response.

Occasionally, NCR fails to reduce aberrant behavior to acceptable levels. This finding has sometimes been reported with cases in which the behavior was maintained by automatic reinforcement. In such cases, NCR can be combined with mild forms of punishment (e.g., response blocking) to further reduce responding to acceptable levels (e.g., Carr, Dozier, Patel, Nicolson, & Martin, 2000; Piazza, Hanley, & Fisher, 1996). Of course, punishment procedures should always be implemented according to the philosophy of the Least Restrictive Alternative (Johnston & Sherman, 1993). However, if punishment becomes necessary in a specific case, NCR is an effective way to maintain the delivery of reinforcers to the individual.

1.5. Behavior-change mechanisms underlying NCR effects

The current understanding in the research literature is that NCR procedures can achieve their effects via three behavior-change mechanisms: (a) attenuation of the reinforcer’s EO, (b) extinction or disruption of the response-reinforcer relationship,

386 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

and (c) reinforcement of other responses that temporally displace the target behavior (Carr, 1996; Fisher et al., 1999). The extinction hypothesis states that, through NCR, the response-reinforcer contingency is eliminated or interrupted, subsequently reduc- ing the aberrant behavior. The success of lean NCR schedules in behavioral reduction supports this position (Hagopian et al., 1994). The EO (or satiation) hypothesis states that the noncontingent delivery of the behavior’s reinforcer eliminates the “motiva- tion” to engage in the behavior. The reductive abilities of dense NCR schedules (e.g., Hagopian et al.) and NCR schedules with high-magnitude reinforcers (e.g., Carr et al., 1998) support this position. Recently, Fisher et al., 1999 reported a study in which a third hypothesis was evaluated. Using within-session analyses, the authors found that, with NCR (without extinction), individuals primarily emitted aberrant behavior during the inter-reinforcement interval. These data suggest that when extinction is not a procedural component of NCR (i.e., the target behavior continues to be reinforced), one putative behavior-change mechanism is the displacement of the target behavior by other responses (possibly during reinforcer “consumption”). In other words, individuals essentially have a choice between “free” and contingent reinforcement. Ultimately, it is likely that NCR achieves its effects through a combination of processes. For example, Vollmer et al. (1993) hypothesized that response reductions during initial continuous NCR schedules were likely a product of satiation and that response reductions observed during schedule thinning were likely products of extinction.

Lalli et al. (1997) provided an interesting method for inferring the behavior- change mechanisms of NCR by analyzing response-reduction patterns. For ex- ample, the authors stated that an immediate reduction in responding is likely an effect of satiation, especially with nearly continuous NCR schedules, as there is little possibility of extinction (see, for example, Hanley, Piazza, Fisher, Con- trucci, & Maglieri, 1997). Similarly, an immediate increase in responding (i.e., a response burst) might be an indication of an extinction process. Increases in responding that are correlated with changes in the NCR schedule (i.e., during schedule thinning) might also be indicative of extinction. Along these same lines, Lattal (1973) and Carr and Britton (1999) have suggested that gradual response reductions (or response maintenance) observed during NCR might be a result of the reinforcer deliveries functioning as discriminative stimuli due to previous contingency experiences. Although there is currently little empiric evidence validating such response-reduction pattern analysis, this preliminary method might prove useful in interpreting future published data sets.

1.6. Directions for future research

It is clear that NCR is an extremely promising technology for the treatment of aberrant behavior. However, the NCR treatment literature is still in its infancy, leaving many research questions unanswered. We propose five topics for future research based on our review of the literature. First, as mentioned above, the exact behavior-change mechanisms by which NCR achieves its effects are not well delineated. Future studies are necessary to determine the

387J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

conditions under which the different mechanisms are involved. Second, adventitious reinforcement remains a potential side effect of NCR. Research is needed to identify the variables responsible for this phenomenon. For example, a controlled comparison of NCR with behaviors occurring at dif- ferent frequencies would determine whether response frequency is a predictor variable for future adventitious reinforcement of the aberrant behavior. Third, although Lalli et al. (1997) provided an effective method for calculating initial NCR schedules, there still is not an agreed upon method for schedule thinning. We expect that this will occur as an ancillary finding within future treatment studies. Fourth, several authors have reported bridge studies that allow NCR to be studied with individuals diagnosed with developmental disabilities using basic-operant preparations (e.g., Carr et al., 1998; Ringdahl & Vollmer, 1998; Wilder et al., in press). This practice has thus far proved successful. Future researchers could use this approach to study topics that are important for the NCR literature, but perhaps not directly so for research participants who exhibit aberrant behavior.

We feel the most important NCR research area that has not yet been adequately explored is the transfer of findings from extremely controlled clinical settings to the natural environment. The vast majority of the literature discussed in this review, involves using NCR to treat aberrant behavior during brief (e.g., 15 min) sessions, with no attention given to extra-session responding. In addition, the terminal thinning schedule for many of the schedules is 5 min. Although a 5-min schedule is much easier to implement than a continuous schedule, it is still impractical for many applied settings (e.g., classrooms). The current NCR treatment literature can perhaps be best described as technology development, as it is extremely difficult to develop and refine new procedures in naturalistic settings. However, the technology must eventually be transferred to the natural environment (e.g., Coleman & Holmes, 1998). It is unclear whether this time has come for NCR procedures; although, we suspect it has. This new line of naturalistic-setting research will probably not be as well controlled as the current studies, but that is to be expected. One of the important outcomes that will result from the transfer is the social validation of NCR procedures, which so far has not been evaluated. Once we complete the transfer of NCR technology from highly controlled (and sometimes analog) settings to more natural- istic ones, NCR will likely be established as one of the most effective treatments for aberrant behavior that has been reported in years.

References

Alleman, H. D., & Zeiler, M. D. (1974). Patterning with fixed-time schedules of response-indepen- dent reinforcement.Journal of the Experimental Analysis of Behavior, 22,135–141.

Boe, R. B. (1977). Economical procedures for the reduction of aggression in a residential setting. Mental Retardation, 15,25–28.

Britton, L. N., Carr, J. E., Kellum, K. K., Dozier, C. L., & Weil, T. M., in press. A variation of noncontingent reinforcement in the treatment of aberrant behavior.Research in Developmental Disabilities.

388 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

Calef, R. S., Choban, M. C., Dickson, M. W., Newman, P. D., Boyle, M., & Baxa, N. D. (1989). The effects of noncontingent reinforcement on the behavior of a previously learned running response. Bulletin of the Psychonomic Society, 27,263–266.

Carr, J. E. (1996). On the use of the term “noncontingent reinforcement.”Journal of Behavior Analysis and Therapy, 1,33–37.

Carr, J. E., B ailey, J. S., Ecott, C. L., Lucker, K. D., & Weil, T. M. (1998). On the effects of noncontingent delivery of differing magnitudes of reinforcement.Journal of Applied Behavior Analysis, 31,313–321.

Carr, J. E., & Britton, L. N. (1999). Idiosyncratic effects of noncontingent reinforcement on problematic speech.Behavioral Interventions, 14,37– 43.

Carr, J. E., Dozier, C. L., Patel, M. R., Nicolson, A. C., & Martin, N. (2000). Separate and combined effects of noncontingent reinforcement and response blocking on object mouthing maintained by automatic reinforcement. Manuscript submitted for publication.

Coleman, C. L., & Holmes, P. A. (1998). The use of noncontingent escape to reduce disruptive behaviors in children with speech delays.Journal of Applied Behavior Analysis, 31,687– 690.

Cooper, J. O., Heron, T. E., & Heward, W. L. (1987).Applied behavior analysis.Columbus, OH: Merrill.

Derby, K. M., Fisher, W. W., & Piazza, C. C. (1996). The effects of contingent and noncontingent attention on self-injury and self-restraint.Journal of Applied Behavior Analysis, 29,107–110.

Dickinson, A., & Charnock, D. J. (1985). Contingency effects with maintained instrumental rein- forcement.The Quarterly Journal of Experimental Psychology, 37B,397– 416.

Edwards, D. D., Peek, V., & Wolfe, F. (1970). Independently delivered food decelerates fixed ratio rates.Journal of the Experimental Analysis of Behavior, 14,301–307.

Fischer, S. M., Iwata, B. A., & Mazaleski, J. L. (1997). Noncontingent delivery of arbitrary reinforcers as treatment for self-injurious behavior.Journal of Applied Behavior Analysis, 30, 239 –249.

Fisher, W. W., Ninness, H. A. C., Piazza, C. C., & Owen-DeSchryver, J. S. (1996). On the reinforcing effects of the content of verbal attention.Journal of Applied Behavior Analysis, 29,235–238.

Fisher, W. W., Thompson, R. H., DeLeon, I. G., Piazza, C. C., Kuhn, D. E., Rodriguez-Catter, V., & Adelinis, J. D. (1999). Noncontingent reinforcement: Effects of satiation versus choice responding.Research in Developmental Disabilities, 20,411– 427.

Goh, H., Iwata, B. A., & DeLeon, I. G. (2000). Competition between noncontingent and contingent reinforcement schedules during response acquisition.Journal of Applied Behavior Analysis, 33, 195–205.

Hagopian, L. P., Fisher, W. W., & Legacy, S. M. (1994). Schedule effects of noncontingent reinforcement on attention-maintained destructive behavior in identical quadruplets.Journal of Applied Behavior Analysis, 27,317–325.

Hagopian, L. P., LeBlanc, L. A., & Maglieri, K. A. (2000). Noncontingent attention for the treatment of excessive medical complaints in a medically fragile man with mental retardation.Research in Developmental Disabilities, 21,215–221.

Halliday, M. S., & Boakes, R. A. (1971). Behavioral contrast and response independent reinforce- ment.Journal of the Experimental Analysis of Behavior, 16,429 – 434.

Hanley, G. P., Piazza, C. C., & Fisher, W. W. (1997a). Noncontingent presentation of attention and alternative stimuli in the treatment of attention-maintained destructive behavior.Journal of Applied Behavior Analysis, 30,229 –237.

Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci, S. A., & Maglieri, K. A. (1997b). Evaluation of client preference for function-based treatment packages.Journal of Applied Behavior Analysis, 30, 459 – 473.

Hart, B. M., Reynolds, N. J., Baer, D. M., Brawley, E. R., & Harris, F. R. (1968). Effects of contingent and noncontingent social reinforcement on the cooperative behavior of a preschool child. Journal of Applied Behavior Analysis, 1,73–76.

Horner, R. D. (1980). The effects of an environmental “enrichment” program on the behavior of institutionalized profoundly retarded children.Journal of Applied Behavior Analysis, 13,473– 491.

389J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1994). Toward a functional analysis of self-injury.Journal of Applied Behavior Analysis, 27,197–209. (Reprinted from Analysis and Intervention in Developmental Disabilities, 2, 3–20, 1982).

Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R., Vollmer, T. R., Smith, R. G., Rodgers, T. A., Lerman, D. C., Shore, B. A., Mazaleski, J. L., Goh, H., Cowdery, G. E., Kalsher, M. J., McCosh, K. C., & Willis, K. D. (1994). The functions of self-injurious behavior: An experimental- epidemiological analysis.Journal of Applied Behavior Analysis, 27,215–240.

Iwata, B. A., Vollmer, T. R., Zarcone, J. R., & Rodgers, T. A. (1993). Treatment classification and selection based on behavioral function. In R. Van Houten & S. Axelrod,Behavior analysis and treatment,(pp. 101–125). New York: Plenum.

Job, R. F. (1988). Interference and facilitation produced by noncontingent reinforcement in the appetitive situation.Animal Learning and Behavior, 16,451– 460.

Johnston, J. M., & Sherman, R. A. (1993). Applying the least restrictive alternative principle to treatment decisions: A legal and behavioral analysis.Behavior Analyst, 16,103–115.

Kahng, S., Iwata, B. A., DeLeon, I. G., & Worsdell, A. S. (1997). Evaluation of the “control over reinforcement” component in functional communication training.Journal of Applied Behavior Analysis, 30,267–277.

Lachter, G. D. (1980). A control procedure in operant conditioning.Perceptual and Motor Skills, 50, 1257–1258.

Lachter, G. D., Cole, B. K., & Schoenfeld, W. N. (1971). Response rate under varying frequency of non-contingent reinforcement.Journal of the Experimental Analysis of Behavior, 15,233–236.

Lalli, J. S., Casey, S. D., & Kates, K. (1997). Noncontingent reinforcement as treatment for severe problem behavior: Some procedural variations.Journal of Applied Behavior Analysis, 30,127–137.

Lalli, J. S., Mace, F. C., Livezey, K., & Kates, K. (1998). Assessment of stimulus generalization gradients in the treatment of self-injurious behavior.Journal of Applied Behavior Analysis, 31, 479 – 483.

Lattal, K. A. (1973). Response-reinforcer dependence and independence in multiple and mixed schedules.Journal of the Experimental Analysis of Behavior, 20,265–271.

Luiselli, J. K. (1994). Effects of noncontingent sensory reinforcement on stereotypic behaviors in a child with posttraumatic neurological impairment.Journal of Behavior Therapy and Experimen- tal Psychiatry, 25,325–330.

Mace, F. C., & Lalli, J. S. (1991). Linking descriptive and experimental analysis in the treatment of bizarre speech.Journal of Applied Behavior Analysis, 24,553–562.

Mace, A. B., Shapiro, E. S., & Mace, F. C. (1998). Effects of warning stimuli for reinforcer withdrawal and task onset on self-injury.Journal of Applied Behavior Analysis, 31,679 – 682.

Marcus, B. A., & Vollmer, T. R. (1996). Combining noncontingent reinforcement and differential reinforcement schedules as treatment for aberrant behavior.Journal of Applied Behavior Anal- ysis, 29,43–51.

Oakes, W. F., Rosenblum, J. L., & Fox, P. E. (1982). “Manna from heaven”: The effect of noncontingent appetitive reinforcers on learning in rats.Bulletin of the Psychonomic Society, 19,123–126.

O’Neill, J. H., & Morris, R. J. (1979). The development of imitation in nonimitative, severely retarded children: Contribution of reinforcement and instruction.Rehabilitation Psychology, 26,79 – 89.

Persel, C. S., Persel, C. H., Ashley, M. J., & Krych, D. K. (1997). The use of noncontingent reinforcement and contingent restraint to reduce physical aggression and self-injurious behaviour in a traumatically brain injured adult.Brain Injury, 11, 751–760.

Piazza, C. C., Contrucci, S. A., Hanley, G. P., & Fisher, W. W. (1997). Nondirective prompting and noncontingent reinforcement in the treatment of destructive behavior during hygiene routines. Journal of Applied Behavior Analysis, 30,705–708.

Piazza, C. C., Fisher, W. W., Hanley, G. P., LeBlanc, L. A., Worsdell, A. S., Lindauer, S. E., & Keeney, K. M. (1998). Treatment of pica through multiple analyses of its reinforcing functions. Journal of Applied Behavior Analysis, 31,165–189.

Piazza, C. C., Hanley, G. P., & Fisher, W. W. (1996). Functional analysis and treatment of cigarette pica. Journal of Applied Behavior Analysis, 29,437– 450.

390 J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391

Poling, A., & Normand, M. (1999). Noncontingent reinforcement: An inappropriate description of time-based schedules that reduce behavior.Journal of Applied Behavior Analysis, 32,237–238.

Rescorla, R. A., & Skucy, J. C. (1969). Effect of response-independent reinforcers during extinction. Journal of Comparative and Physiological Psychology, 67,381–389.

Ringdahl, J. E., & Vollmer, T. R. (1998, May). An evaluation of noncontingent reinforcement effects as a function of baseline reinforcement schedules. In J. J. McComas (Chair),Applications of basic behavioral principles.Symposium conducted at the 24th Annual Convention of the Association for Behavior Analysis, Orlando, FL.

Roscoe, E. M., Iwata, B. A., & Goh, H. (1998). A comparison of noncontingent reinforcement and sensory extinction as treatments for self-injurious behavior.Journal of Applied Behavior Anal- ysis, 31,635– 646.

Sheppard, W. C. (1969). Operant control of infant vocal and motor behavior.Journal of Experimental Child Psychology, 7,36 –51.

Sigafoos, J., & Pennell, D. (1995). Noncontingent application versus contingent removal of tactile stimulation: Effects on self-injury in a young boy with multiple disabilities.Behaviour Change, 12, 139 –143.

Skinner, B. F. (1948). “Superstition” in the pigeon.Journal of Experimental Psychology, 38, 168 –172.

Smith, R. G., Lerman, D. C., & Iwata, B. A. (1996). Self-restraint as positive reinforcement for self-injurious behavior.Journal of Applied Behavior Analysis, 29,99 –102.

Sprague, J., Holland, K., & Thomas, K. (1997). The effect of noncontingent sensory reinforcement, contingent sensory reinforcement, and response interruption on stereotypical and self-injurious behavior.Research in Developmental Disabilities, 18,61–77.

Thelen, M. H. (1979). Treatment of temper tantrum behavior by means of noncontingent positive attention.Journal of Clinical Child Psychology, 8,140.

Tucker, M., Sigafoos, J., & Bushell, H. (1998). Use of noncontingent reinforcement in the treatment of challenging behavior: A review and clinical guide.Behavior Modification, 22,529 –547.

Vollmer, T. R. (1999). Noncontingent reinforcement: Some additional comments.Journal of Applied Behavior Analysis, 32,239 –240.

Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., & Mazaleski, J. L. (1993). The role of attention in the treatment of attention-maintained self-injurious behavior: Noncontingent reinforcement and differential reinforcement of other behavior.Journal of Applied Behavior Analysis, 26,9 –21.

Vollmer, T. R., Marcus, B. A., & Ringdahl, J. E. (1995). Noncontingent escape as treatment for self-injurious behavior maintained by negative reinforcement.Journal of Applied Behavior Analysis, 28,15–26.

Vollmer, T. R., Progar, P. R., Lalli, J. S., Van Camp, C. M., Sierp, B. J., Wright, C. S., Nastasi, J., & Eisenschink, K. J. (1998). Fixed-time schedules attenuate extinction- induced phenomena in the treatment of severe aberrant behavior.Journal of Applied Behavior Analysis, 31,529 –542.

Vollmer, T. R., Ringdahl, J. E., Roane, H. S., & Marcus, B. A. (1997). Negative side effects of noncontingent reinforcement.Journal of Applied Behavior Analysis, 30,161–164.

Wilder, D. A., & Carr, J. E. (1998). Recent advances in the modification of establishing opreaitons to reduce aberrant behavior.Behavioral Interventions, 13,43–59.

Wilder, D. A., & Carr, J. E., & Gaunt, B. T. (in press). On the effects of noncontingent food delivery during naturally occurring periods of deprivation and satiation.Behavioral Interventions.

Wilder, D. A., Draper, R., Williams, W. L., & Higbee, T. S. (1997). A comparison of noncontingent reinforcement, other competing stimulation, and liquid rescheduling for the treatment of rumi- nation.Behavioral Interventions, 12,55– 64.

Zeiler, M. D. (1968). Fixed and variable schedules of response-independent reinforcement.Journal of the Experimental Analysis of Behavior, 11,405– 414.

391J.E. Carr et al. / Research in Developmental Disabilities 21 (2000) 377–391