Professional Documents
Culture Documents
PII S0891-4222(96)00046-7
Jesfis Rosales-Ruiz
University of North Texas
This research fulfilled a requirement for the first author's Master of Arts degree in Human
Development from the University of Kansas.
The authors would like to extend their gratitude to Shahla Alai-Rosales, Donald M. Baer, David
G. Born, Jorge Garcia, Bryan D. Midgley, and James A. Sherman for their contributions in the
development of this project.
Requests for reprints should be sent to Kenda Morrison, Department of Human Development,
University of Kansas, Lawrence, KS 66045-2133.
127
128
Children with autism often display high rates of stereotypic behaviors, such as
hand-flapping, body-rocking, and gazing at lights (Schreibman, 1988). This
finding is clinically significant partially because these behaviors sometimes
interfere with the learning of new behaviors, as well as the performance of those
previously learned (Koegel & Covert, 1972; Lovaas, Litrownik, & Mann, 1971;
Lovaas, Newsom, & Hickman, 1987; Risley, 1968). Due to the apparent
incompatibility of learning and stereotypic behavior, at least in lowerfunctioning children with autism, the reduction or transformation of these
behaviors has been considered necessary for the further development of these
children (Lovaas, 1981).
Stereotypic behaviors have been effectively reduced with response-decrement procedures focusing on the use of consequences. For example, sensory
extinction (Aiken & Salzberg, 1984; Rincover, 1978) and overcorrection (Maag,
Rutherford, Wolchik, & Parks, 1986) have decreased stereotypies.
An alternative strategy is to reduce stereotypy during teaching by removing
the stimulus conditions evoking the stereotypy. This approach was suggested as
early as 1968, when Risley eliminated climbing in a child with autism by
removing the furniture from the room. This strategy, however, can be difficult
because the relevant variables may be difficult to ascertain. For example, high
demand conditions (Mace, Browder, & Lin, 1987), novel therapists (Runco,
Charlop, & Schreibman, 1986), and different types of objects (Watters & Wood,
1983) have all been associated with increased stereotypic behavior. In particular,
LaGrow and Repp (1984) determined that 15 of the 60 studies they reviewed
addressed stereotypic behavior by someone who was manipulating the environment or an object within the environment.
Grandin (1987), who was diagnosed with autism as a child and is now a
successful adult, advises parents and teachers to use their children's fixations as
tools to increase their motivation to learn. Using the objects of children's
fixations as teaching materials is suggested.
There is a growing body of literature that illuminates the importance of
providing choices when working with people with disabilities (Guess, Benson,
& Siegel-Causey, 1985; Koegel et al., 1989; Shevin & Klein, 1984). Providing
choices has been demonstrated to reduce problem behavior (Dunlap et al., 1994;
Dyer, Dunlap, & Winterling, 1990; Foster-Johnson, Ferro, & Dunlap, 1994),
including social avoidance (Koegel, Dyer, & Bell, 1987) and stereotypy (Dyer,
1987). Task engagement (Dunlap et al., 1994), task performance (Mithaug &
Mar, 1980; Parsons, Reid, Reynolds, & Bumgarner, 1990), responsiveness
(Dyer, 1987), and desirable behaviors, such as direction following (FosterJohnson et al., 1994) have increased in association with choice-making. When
the logic of choice is extended to teaching materials, problems could arise if
items associated with high levels of stereotypy, which in turn interfere with
learning, are chosen.
The present study is concerned with both the interfering effects of stereotypy,
as well as the stimulus conditions associated with stereotypy. We analyze the
Preferred Objects
129
Subject
The participant was a 5-year-old child with autism, who had been receiving
intensive behavioral treatment to increase language, social, daily living, recreational, and academic skills for approximately 2 1/2 years. His IQ was rated at
36 by the Bayley Scales of Infant Development and he had minimal functional
language (less than 10 words).
Procedure
Experimental sessions lasted approximately 8 to 10 minutes per day, and
were held up to four times per week.
Counting Task
The counting task consisted of the experimenter asking the child to count a
specified number of objects and place them in a container. At the beginning of
each session, 10 objects of one type (e.g., the 10 bean bags) were placed in front
of the child on the table and an open container was located to the child's right.
If necessary, the experimenter said, "Hands down," and the trial began. The
experimenter then said, "Give me [number]" (e.g., "Give me 5"), and the child
repeated the number (as taught earlier). Then the experimenter said, "Again,"
and the child repeated the number (this verbal rehearsal was used in an attempt
to break echolalic sequences as a part of his overall teaching program). Finally,
the experimenter said, "Give me [number]." The child had been taught previously to count that number of objects into the container, one by one, on this third
request, after two statements of the number. If the child put the correct number
in the container, the experimenter provided approximately 5 seconds of music
from a tape recorder and praised him. If the child was incorrect, the experi-
130
menter said, "no" in a neutral tone of voice. A trial was scored as correct when
the child put in the container only the number of objects that the experimenter
had specified and then put his hands in his lap. An incorrect response was scored
if the child put more or fewer objects in the container than the experimenter had
requested (he often put all the objects in the container), or 15 seconds passed
without him putting any objects in the container.
Behavior Observation
The child's teachers were also the reliability observers. They sat approximately 6 feet from the child and experimenter, and recorded data simultaneously
with the experimenter. Observers recorded the number of items requested, the
number of items given, the type of response (correct, incorrect, or prompted),
absence or presence of stereotypy, eye gaze, and the consequence that occurred
for each trial.
Experimental Design
A multi-element design (cf. Sidman, 1960) was used to evaluate task performance during brief periods with high-, medium-, and low-preference objects.
Each type of object was used in each daily session, but the order in which the
objects were presented each day was chosen at random.
Experimental Phases
Preference was assessed before the experiment began, as well as during the
final phase of the study. The sequence of experimental phases was as follows:
(a) Feedback Alone, (b) Prompting, and (c) Feedback Alone.
Preferred Objects
131
Assessing Preference
Object-preference ratings were determined by presenting the eight different
kinds of objects in two horizontal rows, with 4 groups of like objects per row,
and 10 objects per group. The position of a particular group of objects within
this arrangement was varied systematically over successive presentations, such
that each group of objects appeared in each position one time, but order was
unpredictable. Eight daily presentations were conducted to determine object
preference.
The preference sessions began by seating the child with the objects in front
of him. No instructions were given; the experimenter waited for the child to
choose an object. When one had been chosen, the experimenter removed all
objects of that kind from the table (the group of which the object was a member)
and waited for the child to select another object. This procedure continued until
all objects had been chosen.
The eight sessions determining object preference yielded eight preference
scores (ranks) for each object; the eight ranks per object were then averaged to
produce an average preference rank for each object group. The two object
groups with the greatest variability in their eight session ranks were discarded.
The remaining six object groups were then divided into two sets, A and B, each
containing one high-, one medium-, and one low-preference group (of 10
objects each).
Feedback alone. Sixteen sessions (eight sessions for each of two sets of objects)
were conducted initially to examine the child's counting performance with the
objects. The sets were used on an alternating schedule such that objects in Set
B were used the first day, Set A was used the second day, Set B the third day,
etc., until each set had been used a total of eight times. Thus, on each day, 24
trials (eight trials each with three types of objects; for example, eight trials with
plastic blocks, eight trials with bean bags, eight trials with plastic fruit) were
conducted. During the set of eight trials, each of the numbers from one to eight
was requested one time. The order of specifying the numbers to be counted, as
well as the order in which the object types appeared was randomized. No
prompting or correction occurred during this part of the experiment. Following
a correct response, approximately 5 seconds of a musical tape was played and
the experimenter praised the child. An incorrect response was followed by the
experimenter saying, "no" in a neutral tone of voice.
Prompting sessions. In the next phase of the study, a prompting procedure was
introduced during 24 sessions in an attempt to facilitate performance of the
counting task. Sessions consisted of 10 trials with each of the three object types
of Set A, a total of 30 trials per day. Each 10-trial block now contained 1 to 2
prompted trials. The prompts followed incorrect responses: On these occasions,
the experimenter now put out only the specified number of objects, and counted
132
aloud with the child as he put these objects into the container. In contrast to the
feedback-alone portion of the study, the order of the numbers specified for
counting was not random throughout the session; now, the order of numbers
requested with the first type of object (the first 10 trials) was repeated with the
second and third types of object (the second t0-trial block and third 10-trial
block). The order in which the different groups of objects used was randomized.
Prompts were given in the second and third sets of trials on the occasions yoked
to their use in the first set, whether or not the preceding response was incorrect:
If "three" was prompted on the fifth trial of the first set of trials because the
child had counted it incorrectly, then "three" was prompted on the fifth trials of
the second and third sets of trials as well, whether or not the child had counted
it incorrectly. Except for this, consequences for correct and incorrect responses
were identical to those used in the Feedback Alone condition.
Object Preference
The assessment of object preference ranked bean bags and felt as high
preference, beads and plastic blocks as medium preference, and plastic fruits
and plastic shapes as low preference, puzzle pieces and crayons were discarded
because of the child' s variability in choosing them. The remaining objects were
divided into two sets; Set A consisted of bean bags, plastic blocks, and fruit; Set
B included small felt cutouts, wooden beads, and plastic shapes. When object
preferences were reassessed following the Prompting stage of the experiment,
these rankings proved stable.
Preferred Objects
Stereotypic
Behavior
Feedback
Feedback Alone
Plus
Across
133
Objects
Feedback Alone
Promotlna
100
~'
A 83
~eo
Set_A T
J
I
|~1
, , ,
2 3 4 5 6 7
,
g
[]
.......
I
[] ......
4
6 7
[~.
~.
'd
1 2 3 4 5 6
9 [01112131415161718192021222324
Sessions
Sessions
Sessions
m B0,
G .......
alLe
"6
~40,
Objects
20
1 2 3 4 5 6 7 8
Sessions
iSessions
23456
FIGURE 1. Percentage of correct responses during counting task across object type and
phases.
Feedback Alone phase, 25% of sessions in the Prompting phase, and 17% of
sessions in the Return to Feedback Alone phase. Each interobserver agreement
trial was scored as an agreement or disagreement. Average interobserver agreement on correct and incorrect responses was 100%, stereotypic behavior was
92%, and eye gaze was 97.9%. All disagreements scored regarding eye gaze
occurred when the word "eye" was neither circled nor slashed by the reliability
observer.
134
Correct
S~,A~
't
Responses
Across
Objects
Feedback Alone
Feedback Alone
,'', ~,
g.
2 3 4 5 6 7 5
1 2 3 4 5
Sessions
I ~
~-eo
o.q1",tJ
6 7 8 9101112131415101718192021222324
Seesfons
0 1 2 3 4 5 6
Sessions
HighPreference Objects
, MediumPreference Objects
LOW Preference
Objects
0 2343678
0 1 2 3 4 5 0
Sessions
Sessions
FIGURE 2. Percentage of counting task trials containing stereotypic behavior across object
type and phases.
Stereotypic behavior during the counting task was calculated by dividing the
number of trials during which stereotypy occurred by the total number of trials
with that type of object in each session and multiplying by 100. As shown in
Figure 2, stereotypic behavior occurred more frequently when the child counted
high preference objects (73%), than when he counted medium (37%) or low
preference objects (40%).
Eye gaze was calculated by dividing the number of trials during which eye
gaze was scored as occurring by the number of trials in which the presence or
absence of eye gaze was recorded. Eye gaze occurred during 100% of trials and
across all objects.
DISCUSSION
The present study shows that accurate performance of tasks suffered when
the child engaged in stereotypy. This phenomenon was also observed by Risley
(1968), Koegel and Covert (1972), and Chock and Glahn (t983). Although this
child was more likely to be correct in the absence of stereotypic behavior, there
were occasions when he responded correctly in spite of his ongoing stereotypic
behavior. Thus, stereotypy and correct responding were not entirely incompatible; instead, stereotypy was characteristic of situations in which little learning
occurred.
Preferred Objects
135
Results were consistent with those of other investigators (Mace et al., 1987;
Runco et al., 1986; Watters & Wood, 1983) in suggesting that certain stimuli set the
occasion for higher rates of stereotypy. The highly preferred objects appeared to
evoke stereotypy and thus reduce the possibility of accurate counting.
It is possible that certain object characteristics are associated with higher
levels of stereotypic behavior. Watters and Wood (1983) found that soft toys
evoked more stereotypy than hard or wheeled toys. In this study, too, materials
associated with the highest rates of stereotypy were soft (bean bags and felt).
Conversely, medium- and low-preference materials associated with lower rates
of stereotypy were hard (plastic blocks, plastic shapes, wooden beads, and
plastic fruit). More experience across a wide range of materials and textures may
show how reliably stereotypy can be predicted by material and texture.
Access to stereotypy has been demonstrated to be an effective reinforcer
(Charlop, Kurtz, & Casey, 1990; Hung, 1978; Sugai & White, 1986). Therefore,
it may be difficult for a teacher to provide a stronger extrinsic reinforcer than the
"perceptual reinforcer" produced by using the object for stereotypic purposes
(for related discussion, see Lovaas et al., 1987). The music that was played
following each correct response did not often appear to function as a reinforcer.
Thus, it is possible that engaging in stereotypic behavior with some, but not all,
objects was more reinforcing than listening to music.
Koegel et al. (1989) suggest increasing a child's motivation within a teaching
situation by fostering "shared control." One way to accomplish this is by
permitting the child to choose a toy to play with or task to perform within a
teaching interaction. Indeed, the positive effects of using choices when working
with children with various disabilities has been well documented (Dunlap et al.,
1994; Dyer, 1987; Dyer et al., 1990; Foster-Johnson et al., 1994; Koegel et al.,
1987; Mithaug & Mar, 1980; Parsons et al., 1990). The present study supplements this literature by suggesting one possible parameter to this practice.
It is impossible, however, to ascertain the actual role of object preference on
task performance within this study. The high-preference objects may have been
preferred only because they were more desirable for stereotypic use. This factor
is important because teachers should be aware of this possibility when providing
choices to students. Future research could help to clarify the relevant variables.
Future research needs to be conducted to determine the external validity of
these findings. Would the same results occur: (a) across different (including
more interesting and naturalistic) types of tasks, (b) if the child was demonstrating skill acquisition, (c) if the child had access to the teaching materials
contingent upon a correct performance? Furthermore, only one subject was used
in this study, which raises questions about the generality of these findings.
Although the scope of this research is limited, teachers should be aware that
some objects may produce increased challenges for some children. Teachers
should quantitatively monitor individual student performance, as well as other
behaviors prejudicial to learning, especially when using materials that appear to
evoke stereotypy or are chosen by the child. When these objects are associated
136
with decreased task performance, the teacher may instead present the preferred
items as a reward for completion of a desired task.
REFERENCES
Aiken, J. M., & Salzberg, C. L. (1984). The effects of a sensory extinction procedure on stereotypic
sounds of two autistic children. Journal o/'Autism and Developmental Disorders, 14, 291-299.
Charlop, M. H., Kurtz, P. E, & Casey, E G. (1990). Using aberrant behaviors as reinforcers for
autistic children. Journal of Applied Behavior Analysis, 23, 163-181.
Chock, P. N., & Glahn, T. J. (1983). Learning and self-stimulation in mute and echolalic autistic
children. Jaurnal of Autism and Devehqnnental Disorders, 13, 365-381.
Dunlap, G., DePerczel, M., Clarke, S., Wilson, D., Wright, S., White, R., & Gomez, A. (1994).
Choice making to protnote adaptive behavior tot students with emotional and behavioral
challenges. Journal ~['Applied Behavior Ana(vsis, 27, 5(/5-518.
Dyer, K. (1987). The competition of autistic stereotyped behavior with usual and specially assessed
reinlorcers. Research in Develapmental Disabilities, 8, 607 626.
Dyer, K., Dunlap, G., & Winterling, V. (1990). Effects of choice making on the serious problem
behaviors of students with severe handicaps. Journal ~fApplied Behavior Analysis, 23, 515-524.
Foster-Johnson, L., Ferro, J.. & Dunlap, G. (1994). Preferred curricular activities and reduced
problem behaviors in students with intellectual disabilities. Journal ~fApplied Behavior Anal3,sis, 27, 493-504.
Guess, D., Benson, H. A., & Siegel-Causey, E. (1985). Concepts and issues related to choice-making
and autonomy among persons with severc disabilities. Journal (~( the Association for Persons
with Severe Handicaps, 10. 79-86.
Grandin, T. (1987). Motivating autistic children. Academic Therapy, 22, 297-301.
Hung, D. (1978). Using self-stimulation as reinforcement for autistic children. Journal c?[Autism
and Developmental Disordetw, 8, 355-366.
Koegel, R. L., & Covert, A. (1972). The relationship of serf-stimulation to learning in autistic
children. Journal (~[Apl~lied Behavior Analysis, 5, 381-387.
Koegel, R. L., Dyer, K., & Bell, L. K. (1987). The influence of child-preferred activities on autistic
children's social behavior. Journal (~fAl~plied Behavior Analysis, 20, 243-252.
Koegel, R. L., Schreibman. L., Good, A., Cerniglia, L., Murphy, C., & Koegel, L.K. (1989). How
to teach pivotal behaviors to children with autism: A training manual. Santa Barbara: University
of California.
LaGrow, S. J. & Repp, A. C. (1984). Stercotypic responding: A review of intervention research.
American ,hmrnal of Mental DeJieiency, 88, 595-6(19.
Lovaas, O. I. (1981). Teaching developmentally disabled children: The me book. Austin, Texas:
Pro-ed.
Lovaas, O. l., Litrownik, A., & Mann. R. (1971). Response latencies to auditory stimuli in autistic
children engaged in self-stimulatory behavior. Behaviour Research and Therapy, 9, 305-310.
Lovaas, O. 1.. Newsom, C., & Hickman, C. (1987). Self-stimulatory behavior and perceptual
reinforcement. Journal ~/'Al~plied Behavior Analysis, 20, 4 5 ~ 8 .
Maag, J. W.. Rutherford, R. B., Wolchik, S. A. & Parks, B. T. (1986). Brief report: Comparison of
two short overcorrection procedures on the stereotypic behavior of autistic children. Journal of
Autism and Developmental Disorders, 16, 83-87.
Mace, F. C., Browder, D. M., & Lin, Y. (1987). Analysis of demand conditions associated with
stereotypy. Journal t?f Behal ior Therapy and Experimental Psychiatry, 18, 25-31.
Mithaug, D. E., & Mar, D. K. (1980). The relation between choosing and working prevocational
tasks in two severely retarded young adults. Journal ~fApplied Behavior Analysis, 13, 177-182.
Parsons, M. B., Reid. D. H., Reynolds, J., & Bumgarner, M. (1990). Effects of chosen versus
assigned jobs on the work pertormance of persons with severe handicaps. Journal of Applied
Behavior Analysis, 23, 253-258.
Preferred Objects
137