(solution) NUMBER 1980, 131, 101-117 ESCAPE AS A FACTOR IN THE

(solution) NUMBER 1980, 131, 101-117 ESCAPE AS A FACTOR IN THE

Summarize the article “Escape as a factor in the aggressive behavior of two retarded children”. 

See attachment. 

NUMBER
1980, 131, 101-117
ESCAPE AS A FACTOR IN THE AGGRESSIVE
BEHAVIOR OF TWO RETARDED CHILDREN JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1 (SPRING 1980) EDWARD G. CARR, CRIGHTON D. NEWSOM, AND JODY A. BINKOFF
STATE UNIVERSITY OF NEW YORK AT STONY BROOK AND
SUFFOLK CHILD DEVELOPMENT CENTER This study sought to identify some of the variables controlling the severely aggressive
behavior of two retarded children. In Experiment 1, each child was presented with several demand and nondemand situations. Aggression was frequent in the demand situations and rare in the nondemand situations. When a stimulus correlated with the termination of demands was introduced, aggression fell to a near zero level. In Experiment 2,
for one child, a variety of preferred reinforcers was introduced into the demand situation
contingent on correct responding. Aggression abruptly decreased to a low level. Experiments 3 and 4 involved the second child. In Experiment 3, this child was permitted, in
one condition, to leave the demand situation if he emitted a nonaggressive response.
Aggression decreased to a low level. In Experiment 4, he was prevented, in one condition, from leaving the demand situation in spite of high levels of aggression. Aggression fell to a near zero level. In Experiments 3 and 4, he was permitted, in several
conditions, to leave the demand situation following aggressive behavior. Aggression increased to a high level. The results suggested that: (1) aggression can sometimes function as an escape response; and (2) escape-motivated aggression can be controlled by:
(a) introducing strongly preferred reinforcers to attenuate the aversiveness of the demand situation; (b) strengthening an alternative, nonaggressive escape response; or (c)
using an escape-extinction procedure.
DESCRIPTORS: aggressive behavior, escape, reinforcement, extinction, retarded children No human behavior arouses more social concern than that of aggression. At the level of
the individual child, high-frequency aggressive
behavior often has the effect of disrupting normal family functioning as well as making the
child unteachable in school. Because of these
negative effects, many aggressive children are taken out of the mainstream of society and
placed in institutions. The absence of treatment
in many of these institutions is a particularly
serious shortcoming in the light of evidence
showing that aggressive children often mature
into aggressive adults (Robins, 1966).
Because of the above danger, researchers
have focused much effort on trying to isolate
the factors responsible for the maintenance of
aggression in children. Thus, Patterson, Littman, and Bricker (1967) demonstrated that
the aggressive behaviors of nursery school children generally received high levels of positive
reinforcement from those children against whom
the aggression was directed. The reinforcement
typically took the form of the victim's giving
up his toy or displaying various defensive postures. Reinforcement from adults has also been
implicated as a factor in the maintenance of
aggression in children, particularly when it takes This investigation was supported in part by
U.S.P.H.S. Biomedical Research Support Grant 5 S07
RR07067-11 to S.U.N.Y. at Stony Brook. Portions
of the paper were presented at the annual meeting
of the American Psychological Association, Toronto,
Canada, August 1978, and at the annual meeting of
the Association for the Advancement of Behavior
Therapy, Chicago, Illinois, December 1978. The authors thank their undergraduate assistants, especially
Marti White and Dominic Squittiere, for help with
data collection. We also thank Martin Hamburg, Executive Director, Suffolk Child Development Center,
for his generous support. Requests for reprints should
be sent to Edward Carr, Department of Psychology,
State University of New York, Stony Brook, New
York 11794.
101 EDWARD G. CARR et al. the form of maternal attention delivered contingent on the emission of aggressive acts (Hawkins, Peterson, Schweid, & Bijou, 1966). Restructuring the environment so that aggressive
behavior is no longer followed by the usual
social reinforcers has proven to be a potent
means of controlling this problem behavior
(Brown & Elliott, 1965; Burchard & Tyler,
1695; Hamilton, Stephens, & Allen, 1967;
Wolf, Risley, & Mees, 1964). Finally, vicarious
positive reinforcement of aggressive acts has
also been identified as an important factor in
the maintenance of aggression, affecting the behavior in much the same way as direct reinforcement (Bandura, Ross, & Ross, 1963).
There are indications in the literature that
aggressive behavior may be maintained by factors other than positive reinforcement. In particular, several investigators have implied that
aggression may sometimes function as an escape
response to terminate demands or other aversive
stimuli (Ludwig, Marx, Hill, & Browning,
1969; Patterson, Littman, & Bricker, 1967). To
date, however, there have been no systematic
experimental studies of the possible role of escape factors in the maintenance of aggression.
It would be useful to know what the effects of
such factors might be, both from the standpoint of broadening our understanding of the
variables maintaining aggressive behavior as
well as from the standpoint of clinical treatment.
In the present paper, we present a detailed,
functional analysis of the behavior of two retarded children whose aggression appeared to be
motivated primarily by escape factors. boy had a diagnosis of mental retardation with
autistic features. Both children were untestable
on standard intelligence tests. However, Bob
had a social age of 3.3 yr on the Vineland Social Maturity Scale and Sam scored 2.3 yr on the
same scale. Both children were nonverbal and
understood only simple commands. Neither initiated social interactions but would typically sit
on the floor engaging in self-stimulatory behaviors, especially hand-gazing.
Bob's aggression began when he was 5 yr
old and took the form of scratching, hitting,
kicking, and biting. The behavior occurred with
enough force to draw blood and produce bruises
on his victims, who were almost always adults.
Treatment of his aggression with drugs (Thorazine, Stelazine, Mellaril) was ineffective. Bob
was referred to us by his teachers who complained that he became totally unmanageable
when even a minimal demand, such as having
to sit in a chair, was placed upon him.
Sam's aggression began when he was 6 yr old
and consisted of pinching, hair pulling, and
scratching, again directed almost exclusively at
adults. Several treatments for his aggression,
including drug intervention (Phenerganfortis),
timeout, and the Feingold diet (Feingold, 1975),
had been tried without success. Like Bob, Sam
was referred by his teacher who reported that
all attempts at instruction met with aggressive
behavior.
Setting and recording technique. Sessions with
Bob were conducted 5 days per week in a 3.0by 2.5–m room on his ward. Two chairs were
placed facing each other .4 m apart on one side
of the room, one for Bob and one for the experimenter. The experimenter wore protective
EXPERIMENT 1
clothing, consisting of a thick corduroy coat and
In Experiment 1, we examined the effects of rubber gloves. On the opposite side of the room
a simple demand situation on the aggressive be- were a table and two chairs for the observers.
havior of two retarded children.
The observers used prepared data sheets to tally
the frequency of aggressive responses over 1-min
Method
intervals during each 5-min session. The folSubjects. Bob was a 14-yr-old who lived in lowing behaviors were defined as aggressive:
a state institution. Sam was a 9-yr-old who lived scratching (digging in the fingernails and dragat home and attended a private school. Each ging them across the experimenter's skin or AGGRESSION AS AN ESCAPE BEHAVIOR clothing); hitting (striking the experimenter
with the open or closed hand); kicking (striking
the experimenter with the foot); and biting
(closing the teeth on any part of the experimenter's body).
Sessions with Sam were conducted in a 2.2by 1.5-m training cubicle under conditions identical to Bob's with the following exceptions: (a)
the experimenter did not wear protective clothing because Sam's aggressive responses were not
as damaging as Bob's; (b) sessions were 10 min
long, rather than 5 min, because a longer period
of aggressive behavior was more easily tolerated
in working with Sam; and (c) the following behaviors were defined as aggressive: pinching
(squeezing with a pincer grip), hair pulling
(grasping or pulling the experimenter's hair),
and scratching (defined as above).
The setting and recording technique described
for this experiment was also used for subsequent experiments reported below.
Procedure. The effects of demands were
studied using a reversal design whereby conditions in which demands were presented alternated with conditions in which demands were
withheld. Additionally, during one of the demand conditions, we examined the effects of a
stimulus which reliably signaled the termination
of demands.
Demands. This condition reproduced the
problem situation cited as most typical by each
teacher. For Bob, the situation was being required to sit in a chair; for Sam, it was being
instructed in a buttoning task.
Bob was brought into the room and told to
"Sit down." The experimenter sat down in the
chair in front of him and the session began.
A "sit down" command was given each time
that Bob attempted to leave (i.e., whenever he
raised his buttocks 3 in. (7.5 cm) or more off
the seat of the chair). If necessary, he was physically prompted to sit down again. Three additional commands accompanied by prompts occurred during the first session. No additional
commands or prompts (after the start of a session) were necessary in the following sessions. 103 Thus, the demand consisted of the continuous
requirement that he remain in the chair for the
duration of the session once he was told to sit
down. It must be emphasized that Bob had
never been observed to sit in a chair (without
being coerced) either in class or on the ward;
he always sat on the floor in a corner of the
room. Bob was to be praised at the end of each
period in which he sat for 10 consecutive sec
without aggression. However, because of continuous aggressive behavior, no praise was ever
delivered. Although this made for a somewhat
sterile treatment situation, it was, nonetheless,
the same situation which prevailed daily in his
classroom. Finally, the experimenter signaled
the end of each session by removing his gloves,
after which he led Bob from the room. This
condition was in effect for Bob during sessions
1-3 and 8-17.
Sam was handed a buttoning board at the
start of the session and once every 10 sec thereafter was instructed to button it. Physical
prompts were provided, when necessary, in
order to ensure that a correct response would
occur on virtually every trial. Such prompts
were given if the child failed to comply within
3 sec or if he complied but was unable to force
the button through the hole within 3 sec of the
onset of his response. Each correct response,
whether prompted or unprompted, was reinforced with brief verbal praise because this was
the customary practice in Sam's classroom and
he typically responded by smiling. The experimenter signaled the end of each session by removing the buttoning board and placing it on
a storage shelf, after which he returned Sam
to his regular classroom. This condition was in
effect for Sam during sessions 1-12, 19-24, and 36-43.
No demands. This condition was intended to
approximate those periods of the day when Bob
and Sam were not engaged in any structured
activity. Bob was brought into the room but no
demands were made of him. Invariably, he
would sit on the floor in one corner of the room.
The experimenter, who wore coat and gloves, 104 EDWARD G. CARR et al. maintained a distance of .4 m from Bob, the
distance that would have existed had they
both been seated. This condition was in effect
for Bob during sessions 4-7 and 18-25.
In this condition, Sam was brought into the
room and seated. The experimenter handed Sam
the buttoning board but made no demands. He
maintained a distance of .4 m from the child.
This condition was in effect during sessions
13-18 and 25-35.
Safety signal analysis. During the first imposition of the Demands condition, it became evident that the stimuli used by the experimenter
to end the sessions (i.e., removal of the gloves
in the case of Bob and of the buttoning board
in the case of Sam) had a profound effect on
the rate of aggression of each child. To analyze
this effect further, the following additional manipulations were made during the last Demands
condition with each child. At the end of half of
these sessions with Bob, the experimenter removed his gloves, as he normally did after 5
min, and no longer required Bob to sit in the
chair. Because Bob invariably left his chair at
this point to go to the corner, the experimenter
followed Bob, maintaining a distance of .4 m.
The observer continued to record aggressive acts
for an additional (sixth) min, at the end of
which the experimenter led Bob from the room.
Sessions 8, 10, 11, 13, and 15 were concluded
in the above manner and will be referred to
as Safety Signal sessions (so named because the
removal of the gloves signaled the onset of a
"safety" period during which no more demands
to remain seated were made). At the end of
the other half of the sessions, the experimenter
did not remove his gloves after 5 min and,
again, no longer required Bob to sit in the chair.
Bob invariably remained in his chair, however,
during the minute which followed. At the end
of the sixth min, the experimenter removed
his gloves and led Bob from the room. Sessions
9, 12, 14, 16, and 17 were concluded in this
manner and will be referred to as No Safety
Signal sessions (so named because the gloves
were not removed at the 5 min point). same At the end of half of the sessions of the last
Demands condition with Sam, the experimenter
removed the buttoning board, as he normally
did after the 10 min had elapsed, and stopped
making demands. The observer continued to
record aggressive acts for an additional 5 min,
after which the experimenter led Sam from the
room. Sessions 36, 38, 39, and 40 were concluded in this manner and are therefore labeled
(for the reason just given above) as Safety Signal sessions. At the end of the other half of
the sessions, the experimenter stopped giving
demands but did not remove the buttoning
board. After the 15th min had elapsed, the experimenter removed the buttoning board and
led Sam from the room. Sessions 37, 41, 42,
and 43 were concluded in this way and are labeled as No Safety Signal sessions. It should be
noted that Sam was intentionally exposed to
the additional 5-min period at the end of the
regular demand sessions, rather than the 1-min
period used for Bob, in order to determine
whether the effects observed for Bob could also
be obtained with the longer time period. The
procedure therefore constitutes an attempt at
systematic replication (Sidman, 1960).
Reliability was assessed 11 times for Bob
and 9 times for Sam, at least one assessment
being made in each experimental condition. In
this and all subsequent experiments, three reliability observers, each naive to the purpose of
the experiment, were randomly assigned in pairs
to record in various sessions. The response definitions were described verbally to each observer
before the session began. During each session,
the experimenter and two observers were present. For protection, the observers were seated
behind a table .8 m wide, positioned a minimum
of .9 m from the child. During the session, one
observer signaled the end of each 1-min interval
by tapping the leg of the other observer. Minuteby-minute monitoring was used for convenience
because later analyses (e.g., Bob's performance
in the Safety Signal condition) required frequency counts for each minute. The reliability
index was the percentage of agreement between 105 AGGRESSION AS AN ESCAPE BEHAVIOR the two observers, calculated separately for each Demands condition, the mean frequency of agsession by dividing the smaller total frequency gressive acts was 121.7 per 5-min session. Howby the larger. The median interobserver reliabil- ever, with the introduction of the No Demands
ity with Bob was 98% (range: 86%o to 100%), condition in session 4, the rate abruptly fell to
and with Sam, it was 93% (range: 78% to zero. When Demands were reinstated, the mean
100%).
rate rose sharply to 128.3, falling again to .8
during the final No Demands condition.
Results and Discussion
The bottom half of Figure 1 shows the numThe top half of Figure 1 shows the number ber of aggressive responses in each 10-min sesof aggressive responses in each 5-min session sion for Sam. Once again, the frequency of
for the two experimental conditions for Bob. aggressive acts was high during the Demands
The frequency of aggressive acts was high dur- condition and at or near zero in the No Deing the Demands condition and near zero dur- mands condition. During the first, second, and
ing the No Demands condition. During the first third Demands condition, the mean frequency of BOB 1160180 -W II
c CO, ,0 40 2.00 40 QX 0 '4' '4
40 IT 140 – I-a Cl) w 120 0 Cn) z 0
a.
C,) w
w 100 80
60'
40 C,)
cn
w CD 20
0 4 SAM LL 00 NO DEMANDS DEMANDS DEMANDS w 20 z IZ NO DEMANDS DEMANDS 00- 80- 640-0- 20,
O 6 I 0-00-19-10- I
I 5 10 15 I
I 20 ;== 25 30 35 A 4% 40 SESSIONS
Fig. 1. Number of aggressive responses exhibited by Bob and Sam in each session of the Demands and
No Demands conditions. 106 EDWARD G. CARR et al. BOB
NO SAFETY SIGNAL SAFETY SIGNAL
C,)
w
C,) 50
40 z 0 30 w
w 20- >:
wb
0
(D) 00 aC() 0
4L
1La mommolmlov 10
a-r-& 5 6 5 SAM
100. w 80
so W 60 0
a 40
D 20 0
6-10 "1-'5 6-10 I"-'5 TIME PERIOD IN SESSION (min)
Fig. 2. Number of aggressive responses during the 5th and 6th min of each session in the second Demands
condition for Bob (top half of figure) and during the 6th to 10th min and 11th to 15th min of each session in
the third Demands condition for Sam (bottom half of figure). The left panels show data taken during Safety
Signal sessions in which the experimenter signaled to each child that demands were no longer forthcoming. The signal was given to Bob at the start of the 6th min and to Sam at the start of the 11th min. The right
panels show data taken during No Safety Signal sessions in which neither child received any cue that demands had ended. Some of the data points have been slightly displaced horizontally to enhance presentation clarity. 43.3, 53.7, and 91.1, respecThe top half of Figure 2 shows the number
tively. During the first and second No Demands of aggressive responses during the fifth and sixth
condition, the mean frequency of aggressive min of each of the 10 sessions in the second
acts was 1.0 and .3, respectively.
Demands condition for Bob. The left panel
aggressive acts was AGGRESSION AS AN ESCAPE BEHAVIOR shows data taken during the five Safety Signal
sessions and the right panel, data taken during
the five No Safety Signal sessions. During the
fifth min of the Safety Signal sessions, the mean
number of aggressive acts was 26.6. However,
during the sixth min, after the experimenter had
signaled the termination of demands by removing his gloves, the mean number of aggressive
acts fell to .4. In contrast, during the No Safety
Signal sessions, the mean number of aggressive
acts increased from a level of 24.6 during the
fifth min to 30.6 during the sixth min.
The bottom half of Figure 2 shows the number of aggressive responses during the last 5
min of the regular session (i.e., the sixth to
tenth min) and the 5-min period which followed the end of the regular session (i.e., the
11th to 15th min) for each of the 8 sessions
in the third Demands condition for Sam. The
left panel shows the data for the Safety Signal
sessions, in which the number of aggressive acts
fell from a mean of 56.5 during the last 5 min
of the session to a mean of 3.0 during the 11th
to 15th min, the time period which followed the
experimenter's signal that demands had ended.
During the No Safety Signal sessions (right
panel), the number of aggressive acts increased
from a mean of 55.8 during the last 5 min to
a mean of 74.8 during the additional 5 min.
The high rate of aggression which occurred
in the No Safety Signal condition was somewhat
paradoxical since this condition was identical
to the No Demands condition which produced
almost no aggression. One explanation for this
finding is that the absence of a safety signal
during the No Safety Signal condition made
it extremely difficult for the child to discriminate
that the demands had indeed ended and therefore he responded as if the demands condition
were still in effect, namely by continuing to
exhibit aggressive behaviors. In contrast, in the
No Demands condition, aggression was not set
off in the first place since the child was never
given any demands. In short, in the No Demands condition, the child was not presented
with any stimulus that set off aggression, and 107 in the No Safety Signal condition the child was
not presented with any stimulus that stopped
aggression.
Experiment 1 demonstrated that a situation
in which demands occur can be discriminative
for high rates of aggressive behavior. Second,
aggressive responding can be virtually eliminated when a stimulus that is correlated with
the termination of demands is introduced. A
plausible interpretation of the above results is
that aggressive behavior can, under some circumstances, be conceptualized as an escape response. This interpretation draws its strongest
support from the observation that Bob's rate of
aggression dropped dramatically when the experimenter removed his gloves, and Sam's aggression abruptly decreased after the experimenter removed the buttoning board. These
events were always used to signal the end of a
session and were thus highly discriminative for
the termination of demands. A stimulus which
consistently signals the absence of an aversive
event (such as demands might be) is technically
referred to as a "safety signal" (Mowrer, 1960,
p. 129). Typically, operant escape responding
decreases in the presence of a safety signal (Azrin, Hake, Holz, & Hutchinson, 1965, p. 39).
The fact that Bob and Sam stopped aggressing
whenever the experimenter removed his gloves
or the buttoning board suggests that these stimuli functioned as safety signals for them, indicating that the demands had ended and no
further escape responses (aggressive acts) were
necessary. In contrast, it is important to note
that when the experimenter did not remove his
gloves or the buttoning board (No Safety Signal
sessions), thereby failing to signal that demands
had terminated, Bob and Sam continued to aggress against the experimenter. EXPERIMENT 2
The results of Experiment 1 suggested that
aggression may function as an escape-motivated
response that serves to terminate an aversive
demand situation. If this interpretation is valid, EDWARD G. CARR et al. then any treatment intervention aimed at making the demand situation less aversive should
decrease the frequency of aggressive behavior.
One plausible procedure for mitigating the aversiveness of the demand situation would be to
introduce a variety of positive reinforcers known
to be strongly preferred by the child. Accordingly, in the experiment that follows, we attempted to determine the effect of including
such positive reinforcers in the demand sessions
on the level of aggressive behavior exhibited in
those sessions. Method
Sam alone participated because, as described
below, a separate set of procedures was required
to control Bob's aggression.
Procedure. The relationship between added
positive reinforcement (in the form of toys and
food) for correct responding, and the frequency
of aggressive behavior was studied in a reversal
design. ("Correct responding" refers to all correct responses, whether prompted or unprompted.) A condition in which brief verbal
praise followed each correct response (the Demands condition) was alternated with a condition in which correct responses were followed
by toy and food reinforcers in addition to praise
(the Demands Plus Toys and Food condition).
Demands. This condition was identical to the
Demands condition of Experiment 1 for Sam.
It was in effect during sessions 1-5 and 13-15.
Demands plus toys and food. This condition
was identical to the Demands condition, except
that the experimenter now dispensed one of
several toy or food reinforcers (in addition to
the customary verbal praise) each time that
Sam made a correct response. The reinforcers
used were selected on the basis of a two-step
process. In Step 1, the teacher was interviewed
and a list of potential reinfor…