See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/222313877

A variation of noncontingent reinforcement in the

treatment of aberrant behavior

Article  in  Research in Developmental Disabilities · November 2000

DOI: 10.1016/S0891-4222(00)00056-1

CITATIONS READS

35 734

5 authors, including:

Lisa Britton James Edward Carr

Chicago School of Professional Psychology Behavior Analyst Certification Board
14 PUBLICATIONS   470 CITATIONS    182 PUBLICATIONS   4,792 CITATIONS   

SEE PROFILE SEE PROFILE

Karen Kate Kellum Claudia L Dozier

University of Mississippi University of Kansas
28 PUBLICATIONS   602 CITATIONS    50 PUBLICATIONS   1,277 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Representation of Women in Behavior Analysis: An Empirical Investigation View project

RECALL Reading Intervention for Preschool and Kindergarten Age Children with Moderate to Severe Language Delays
View project

All content following this page was uploaded by James Edward Carr on 01 February 2018.

The user has requested enhancement of the downloaded file.

 Research in Developmental Disabilities
21 (2000) 425– 435

A variation of noncontingent reinforcement in

the treatment of aberrant behavior
Lisa N. Britton,1 James E. Carr, Karen Kate Kellum,
Claudia L. Dozier, Timothy M. Weil
University of Nevada, Reno, Nevada

Abstract
We examined the effectiveness of a variation of noncontingent reinforcement (NCR)
that incorporated a stimulus-delay procedure in the reduction of aberrant behavior main-
tained by positive reinforcement. Functional analyses for three individuals diagnosed with
developmental disabilities indicated that their behaviors were maintained by positive
reinforcement; one in the form of access to a tangible item, another by attention, and the
third by physical contact. We implemented NCR with the delay procedure with two
participants using reversal designs to evaluate effects. We also compared this NCR
variation and DRO with the third participant to evaluate reinforcer-delivery rates. The
variation of NCR was successful in reducing all aberrant behavior to near-zero levels. A
comparison of reinforcer delivery between NCR with the stimulus-delay procedure and
DRO demonstrated that the participant accessed more reinforcement with NCR. Results
are discussed in the context of enhancing decelerative interventions with emphases on
minimizing response effort for caregivers and maximizing access to reinforcement for the
individuals. © 2000 Elsevier Science Ltd. All rights reserved.

1. Introduction

Functional analysis refers to the experimental manipulation of antecedents and

consequences for the purpose of identifying variables that maintain aberrant
behavior. Using treatments prescribed from functional analyses to reduce aber-

* Corresponding author. Tel: ⫹1-510-845-1321.

E-mail address: lbritton@spectrumcenter.org (L.N. Britton).
1
Lisa N. Britton’s current address is: Spectrum Center, 2855 Telegraph Ave., Suite 312,
Berkeley, CA 94705.

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 6 - 1
426 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435

rant behavior has become common practice in recent years (e.g., Vollmer,
Marcus, & Ringdahl, 1995). One function-based intervention that has recently
received much research attention is noncontingent reinforcement (NCR). NCR is
the delivery of a behavior’s known reinforcer on a response-independent schedule.
Vollmer, Iwata, Zarcone, Smith, and Mazaleski (1993) compared the effects
of NCR to differential reinforcement of zero rates of responding (DRO) with
three women diagnosed with mental retardation who exhibited self-injurious
behavior (SIB). Both procedures effectively reduced SIB. Based on their results,
the authors proposed several benefits of NCR, including (a) a lower probability
of extinction-induced responding, (b) a higher rate of reinforcer delivery com-
pared to DRO, and (c) relative ease of implementation.
One potential limitation of NCR is that the reinforcer may occasionally follow
the target behavior, resulting in adventitious reinforcement. An initial, dense
schedule of reinforcement that exceeds the rate of the aberrant behavior may
reduce the potential for this problem (Hagopian, Fisher, & Legacy, 1994).
Vollmer, Ringdahl, Roane, and Marcus (1997) documented the first reported case
of negative side effects associated with NCR. The authors found that NCR
resulted in adventitious reinforcement during the treatment of aggression in a girl
diagnosed with mental retardation. The authors presented within-session data indi-
cating that responding increased immediately before the scheduled reinforcer deliv-
ery, indicating that the individual was responding on a fixed-interval schedule of
reinforcement. The authors then implemented a momentary DRO procedure that
eliminated the adventitious contingency, which resulted in a decrease in aggression.
One way to prevent adventitious reinforcement from occurring, while still
using a treatment that is relatively easy to implement, is the use of NCR with a
stimulus-delay feature. That is, if the target behavior occurs in close temporal
proximity to the scheduled delivery, the therapist would wait until the individual
did not engage in the target behavior for a specified amount of time (e.g., 10 s)
before delivering the reinforcer. The modified NCR procedure could be pro-
grammed on a dense schedule (i.e., maximum access to reinforcement) and still
prevent adventitious reinforcement. Lalli, Mace, Livezey, and Kates (1998)
effectively used this procedure to reduce SIB in a single case. The purpose of the
current study was to extend the research on NCR by using a fixed-time (FT)
schedule with a stimulus-delay feature in the treatment of aberrant behavior in
three participants diagnosed with mental retardation. A secondary goal was to
compare the effects of the modified NCR procedure with DRO on rate of
reinforcer delivery.

2. General method

2.1. Participants and setting

Three participants were included. Rob was a 23-year-old male diagnosed with
severe mental retardation who received services from a university-sponsored day
 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435 427

program. At the time of the study, Rob could emit three manual signs, which
were used to request tangible items. He frequently engaged in tantrumming
(flopping to the ground). Todd was a 7-year-old male diagnosed with autism.
Todd was enrolled in a university-sponsored clinic that provided intensive
behavioral treatment for children with developmental disabilities. Todd could not
emit any verbal responses. Todd’s aberrant behavior was inserting his finger into
his nasal cavity. His mother was concerned about the social implications of this
behavior as Todd began to be integrated into a classroom setting. Victor was a
41-year-old male diagnosed with profound mental retardation who was receiving
consultation services from a university-based program. Victor could gesture for
things he wanted. Victor’s aberrant behavior was aggression, including grabbing,
hair pulling, and biting.
For Rob, sessions were conducted in a 1.3 m by 1.3 m cubicle that was located
next to his training area. Todd’s sessions were conducted in the room (4 m by
4 m) used for his regular training sessions. Victor’s sessions were conducted in
his bedroom (3 m by 3 m) at the community-based group home in which he
resided.

2.1.1. Data collection: dependent variable

Rob’s tantrumming was defined as any time one or both knees touched the
floor. Frequency data were collected on the tantrumming due to the discrete
nature of the response. We defined Todd’s finger insertion into his nasal cavity
as any time a finger touched the inside of his nose. Data were collected on nasal
finger insertion using a partial-interval 10-s recording system due to the high
frequency and variable duration of the response. We defined Victor’s aggression
as any time he touched any part of the therapist (except for his/her hand) with his
hand or mouth. Data for aggression were collected using a partial-interval 10-s
recording system due to the variable duration of the responses. All functional
analysis and treatment evaluation sessions were 10 min in length. For each
participant, we conducted approximately 16 sessions per week.
Interobserver agreement (IOA) was calculated for an average of 27% (range,
25% to 29%) of sessions equally distributed across participants and conditions.
IOA was collected for Rob’s data by dividing the lower frequency by the higher
frequency and multiplying by 100%. Mean IOA for Rob’s functional analysis
and treatment evaluation sessions was 100%. IOA was calculated for Todd’s
nasal finger insertion and Victor’s aggression by dividing the number of agree-
ments by the number of agreements plus disagreements and multiplying by
100%. Mean IOA for Todd’s functional analysis data was 95% (range, 83% to
100%) and the mean IOA for Todd’s treatment evaluation sessions was 94%
(range, 80% to 100%). The mean IOA for Victor’s functional analysis data was
98% (range, 90% to 100%) and the mean IOA for his treatment evaluation data
was 98% (range, 90% to 100%). Independent variable integrity data were also
collected and are available from the authors upon request.
428 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435

2.2. Phase 1: functional analysis

2.2.1. Method
Participants’ aberrant behaviors were initially assessed via functional analyses
(Vollmer, Marcus, Ringdahl, & Roane, 1995). Conditions included attention,
tangible, task escape, physical contact, social escape, no interaction, and control.
We used the no-interaction condition instead of an alone condition for Rob and
Todd because the analysis was conducted in the work environment, an environ-
ment in which they were never alone. We used this condition for Victor because
of the nature of the target response. During the physical contact, the therapist did
not provide vocal attention and during the attention condition, the therapist did
not provide physical contact. Test conditions were determined by staff report. We
used a multielement design to evaluate behavioral function. The consequences
were provided on continuous schedules during each session. We conducted the
sessions on a semi-random schedule with a no-interaction phase never following
an attention phase to reduce the likelihood of an extinction burst for an attention-
maintained behavior.

2.2.2. Results and discussion

Figure 1 shows functional analysis data for the three participants. Rob’s
functional analysis indicated that tantrumming was maintained by access to
tangible items. The tangible item used in these sessions was a bucket containing
brightly-colored blocks.
For Todd’s functional analysis, the sessions displayed on the graph are not
continuous. Initially, sessions were conducted with Todd sitting in a chair with
the therapist. There were near-zero levels of responding across sessions during
this assessment. We began conducting sessions with Todd walking around with
the therapist because the direct-service staff indicated that they only observed the
behavior in the natural environment during transitions. We were unable to
replicate the specific response patterns observed during the in-chair analysis,
thus, these data are not represented on the figure. However, these data are
available from the authors upon request. We observed a clear differentiation such
that attention appeared to be maintaining the behavior. However, responding
began to increase over time in the task escape condition. There are two possible
reasons for the increase in responding during this condition. First, Todd may not
have been able to discriminate the task escape condition contingency from the
other conditions initially during the functional analysis. Second, it is possible that
Todd’s behavior could have acquired a new function as a result of the functional
analysis conditions. We used attention to evaluate the treatment because it
appeared to occur at significantly higher intervals than the task escape condition
indicating that it was the primary function of the target behavior. The treatment
evaluation was conducted with Todd and his therapist walking around the room
in a manner similar to that observed in the functional analysis conditions.
Although there was high responding in several of the conditions during the
initial sessions of Victor’s functional analysis, we observed differentiation over
 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435 429

Fig. 1. Frequency per minute of tantrumming during functional analysis conditions for Rob (top
panel). Percentage occurrence of nasal finger insertion during functional analysis conditions for Todd
(middle panel). Percentage occurrence of aggression during functional analysis conditions for Victor
(bottom panel).

time. This differentiation indicated that aggression was maintained by access to

physical contact, which occurred in the form of handholding. Therefore, access
to handholding was most likely the maintaining variable.
430 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435

2.3. Phase 2: treatment evaluation

2.3.1. Method
2.3.1.1. Procedures and experimental design
Baseline. We used an analog baseline to establish the level of behavior before
implementing treatment. For Rob, the therapist provided access to the bucket each
time he fell to the floor. The initial baseline sessions (1 through 5) for Rob were the
tangible sessions from his functional analysis. The initial baseline sessions for Todd
were the last 5 attention sessions from the functional analysis. The baseline sessions
for Victor were all of the physical contact sessions from the functional analysis.
Treatment. We based the initial schedule of reinforcement on the mean la-
tency to the first occurrence of the aberrant behavior during baseline (Lalli,
Casey, & Kates, 1997). For Rob, treatment consisted of providing access to the
bucket on an FT 45-s schedule. For Todd, treatment consisted of providing
attention on an FT 100-s schedule. In all cases, the therapist waited 10 s to
provide the reinforcer if the therapist observed the target behavior within 10 s of
the time of reinforcer delivery. If the participant was still engaging in the aberrant
behavior, the therapist continued to wait 10 s to provide the reinforcer. The next
schedule component did not begin until the therapist delivered that reinforcer.
The individual received access to the reinforcer for 20 s each time it was
delivered. We thinned the schedule once a reduction in the target behavior was
observed. The criterion for thinning the schedule for Rob was three consecutive
sessions of responding less than 1 response per minute and for Todd it was 3 sessions
in a row of responding less than 5% of intervals. We thinned the schedule from 100 s
to 2 min and after that point, the schedule was thinned in 1-min increments. An
ABAB reversal design was used to evaluate intervention effects.
For Victor, a multielement design was used to compare the modified NCR
procedure (as described above) with whole-interval DRO using a nonresetting
interval. During the DRO procedure, Victor received the reinforcer only when
the previous interval contained no instances of the target behavior. Victor
received access to handholding for 20 s during each reinforcer delivery for both
treatments. One therapist conducted the NCR sessions while another therapist
conducted the DRO sessions in an effort to reduce the likelihood of carryover
effects. Once we observed a reduction in both treatments, a reversal design was
implemented. We thinned both of the schedules once we observed a reduction in
the target behavior. The criterion for thinning was 3 consecutive sessions of
responding less than 5% of the intervals. We thinned Victor’s schedules by
increasing the schedule by 50% of the previous schedule.

3. Results

Treatment data for each participant are shown in Figure 2. There was an
increasing trend during Rob’s baseline sessions. We observed a clear reduction
 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435 431

Fig. 2. Frequency per minute of tantrumming during baseline and NCR conditions during the
treatment evaluation for Rob (top panel). Percentage occurrence of nasal finger insertion during
baseline and NCR conditions during the treatment evaluation for Todd (middle panel). Percentage
occurrence of aggression during baseline, NCR, and DRO during the treatment evaluation for Victor
(bottom panel).

in the target behavior when the therapist implemented treatment. We conducted

a reversal to observe the tantrumming, which occurred at rates higher than
baseline. When the therapist implemented treatment a second time, tantrumming
432 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435

was reduced to low levels again. Responding remained low as we thinned the
schedule to 5 min. When we first implemented treatment, Rob began engaging in
head banging at a high rate. The head banging was a behavior that was observed
in the natural environment when Rob was denied access to preferred stimuli.
Since a functional analysis was not conducted on this response prior to treatment
evaluation, these data were not included in this analysis. However, these data are
available from the authors upon request. Rob’s head banging was addressed
clinically following the study.
Todd’s baseline produced high and variable responding. There was a clear
reduction in nasal finger insertion during treatment. Data in the reversal phase
were similar to baseline. The return to treatment was marked by another sharp
decrease in the percent of intervals in which he engaged in the target behavior.
When we thinned the schedule to 2 min, we observed an increase in the behavior.
The criterion for returning to the previous schedule was 3 consecutive sessions
of responding above 5% of the intervals. Responding reached criterion in the
34th session so we returned the schedule to 100 s. There were then 3 consecutive
sessions of responding below 5%, so we increased the schedule again. We
increased the schedule to 5 min while maintaining the reduction in the target
behavior.
We observed an increasing trend during Victor’s baseline. A clear reduction
in aggression was observed for both treatments. During the reversal phase,
responding increased to levels similar to that observed at baseline. We observed
an initial increase in responding when we reintroduced the treatments, which was
followed by decreases in responding. The increases in responding we observed
during thinning of the treatment schedules corresponded to the first session
conducted on each day the therapist implemented treatment.
For each session that we observed responding, we divided the number of times
the therapist delivered the reinforcer by the total number of possible reinforcer
deliveries. In this way, we were able to determine if Victor received more
reinforcement under one treatment over the other. During the 50-s schedule,
Victor received reinforcement for 72% of the opportunities possible during DRO
and for 90% of the opportunities possible during NCR. During the 75-s schedule,
Victor received reinforcement for 88% of the opportunities possible during DRO
and for 89% of the opportunities possible during NCR. For the 113-s schedule,
Victor received reinforcement for 56% of the opportunities possible during DRO
and for 100% of the opportunities possible during NCR. There were no differ-
ences in the final schedules; responding was too infrequent to affect delivery.

4. General discussion

We observed clear reductions in the target behaviors for all participants as a

result of the modified NCR procedure. Rob’s functional analysis results indicated
that his tantrumming was maintained by access to tangible items. The treatment
 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435 433

data indicated that NCR with a stimulus delay was effective in reducing the target
behavior, even when the schedule was thinned to 5 min. Todd’s functional
analysis indicated that the behavior of inserting his finger into his nasal cavity
was maintained by attention and possibly escape. The results of the treatment
evaluation indicated that the modified NCR procedure was effective in reducing
the target behavior. Victor’s functional analysis indicated that aggression was
maintained by access to physical contact. The results of the treatment com-
parison indicate that both NCR and DRO were effective in reducing the target
behavior. The reinforcer delivery data demonstrated that Victor received
more reinforcement during the NCR with delay when he was engaging in
aggression.
The behavior-change mechanisms responsible for the effects observed in the
modified NCR conditions appear to be satiation and extinction. Lalli et al. (1997)
discussed how the response patterns observed during the thinning of NCR
schedules might indicate the operative mechanism. In Rob’s treatment data, high
rates of tantrumming continued when the therapist initially implemented treat-
ment. A reduction in the target behavior followed this increase, indicating that
extinction may have been responsible for the reduction. In Todd’s treatment data,
several bursts in responding followed an initial reduction in the target behavior.
This finding is indicative of satiation since this behavior-change mechanism was
initially responsible for the reduction. Extinction have been responsible for
maintaining the reduction as we thinned the schedule. The initial increase in
responding observed when treatment was implemented for Victor indicates that
extinction might have been responsible for the reduction observed. The fact that
we observed increases for the first session conducted on a particular day also
lends support for this claim.
Although we observed similar reductions in aggression with both NCR and
DRO in Victor’s case, the advantage of using the modified NCR procedure was
ease of implementation. Caregivers are not required to observe the client at all
times to determine if they should deliver the reinforcer. The staff members are
only required to determine if the individual is engaging in the behavior within
10 s of reinforcer delivery. Another advantage of the modified NCR procedure is
that the individual may receive a higher rate of reinforcement compared to
DRO, especially if the therapists continue to observe responding once they
implement treatment. If the rate of reinforcer delivery is higher, extinction
bursts and emotional behavior associated with extinction should be less likely
to occur compared to DRO. Although the contingency between the target
behavior and access to reinforcement is terminated during NCR, the higher
rate of reinforcer delivery would be more likely to attenuate deprivation. In
Victor’s case, bursts in responding did occur. However, we only observed
these bursts for one session at a time and near-zero levels of responding
followed them.
A limitation of the current study is that the modified NCR procedure was not
compared with a traditional NCR procedure. An advantage of NCR with a delay
feature over traditional NCR is that there is no risk of adventitiously reinforcing
434 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435

the target behavior. This issue is especially relevant with high-rate behaviors,
which are more likely to occur around the time of reinforcer delivery. In the
future researchers might directly compare these two procedures with respect to
issues of contiguity. This comparison is relevant because the traditional NCR
procedure is easier to implement as no judgement has to be made on the
occurrence of aberrant behavior. In other words, direct-service staff do not need
to observe for the behavior before delivering the reinforcer with NCR. With this
relative ease of implementation when compared to NCR with a delay feature, and
the low risk of adventitious reinforcement when examining a low frequency
behavior, traditional NCR may be the treatment of choice under those conditions.
Thus far, only one documented case of adventitious reinforcement as a result of
using NCR has been reported (Vollmer et al., 1997). Researchers might collect
data on the prevalence of adventitious reinforcement when using NCR to deter-
mine when NCR with a delay feature is a necessary treatment alternative. If we
can determine certain variables that may be present that make adventitious
reinforcement more likely (e.g. high-frequency behaviors), we can make predic-
tions as to when this modified NCR procedure should be implemented.
Future research areas also include examining social validity measures to
determine which treatments commonly used to decrease aberrant behavior are
preferred by direct-service staff. These staff members are more likely to imple-
ment treatments that are relatively easy to implement in different contexts.
Research on a variety of different techniques is necessary; however, if they are
not used in the natural environment, their utility is limited.
Although we have described the modified NCR procedure as “NCR with a
stimulus-delay” feature, the schedule could also be technically described a
conjunctive, or possibly tandem schedule with FI (DRO) and FT (NCR) com-
ponents. However, regardless of its technical description, it appears to be an
effective variation of the NCR “family” of interventions that retains the
relative ease of implementation and high rate of reinforcer delivery, while
eliminating adventitious reinforcement, perhaps the most troublesome side
effect of NCR.

Acknowledgment

This study is based on a thesis submitted by the first author, under the
supervision of the second author, to the Department of Psychology at University
of Nevada for the MA degree. Lisa Britton is now at Spectrum Center. James
Carr is now at Western Michigan University. Claudia Dozier is now at The
University of Florida. We thank Larry Williams and Mary Anne Demchak for
their helpful comments on an earlier version of the manuscript and Jamie Cuva,
Kimberley Enloe, Heidi Landaburu, Michiko Okura, Ryan Pabico, Jill Romick,
and Kimberlee Romick for their assistance with data collection.
 L.N. Britton et al. / Research in Developmental Disabilities 21 (2000) 425– 435 435

References

Hagopian, L. R., 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.
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.
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 rein-
forcement 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., Marcus, B. A., Ringdahl, J. E., & Roane, H. S. (1995). Progressing from brief
assessments to extended experimental analyses in the evaluation of aberrant behavior. Journal of
Applied Behavior Analysis, 28, 561–576.
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.

View publication stats