443981

Taylor and DeQuinzioBehavior Modification

BMO36310.1177/0145445512443981

Behavior Modification

Observational Learning 36(3) 341­–360

© The Author(s) 2012
Reprints and permission:
and Children With Autism sagepub.com/journalsPermissions.nav
DOI: 10.1177/0145445512443981
http://bmo.sagepub.com

Bridget A.Taylor1 and Jaime A. DeQuinzio1

Abstract
A skill essential for successful inclusion in general education settings is the
ability to learn by observing others. Research, however, has documented chil-
dren with autism display significant deficits in the fundamental skills neces-
sary for observational learning. This article outlines the skills essential for
observational learning from an operant learning perspective, the research
base on teaching observational learning to children with autism, and suggests
practical strategies to increase these skills in children with autism so they
may more fully benefit from inclusion in general education settings.

Keywords
observational learning, autism, inclusion

Learning by observing others has significant educational, economic, and

social implications. If one can learn by observing the consequences deliv-
ered to another, it can reduce instructional time and financial costs related
to intensive instruction, and lead to the acquisition of socially relevant
behavior, thereby increasing social integration opportunities (Ledford,
Gast, Luscre, & Ayres, 2008). Research has demonstrated, however, that
children with autism display deficits in core skills required for observational
learning. Children with autism, for example, display deficits in attending
(Patten & Watson, 2011; Senju, Yaguchi, Tojo, & Hasegawa, 2003), imitat-
ing (Smith & Bryson, 1994), and in making simple and complex auditory

1
Alpine Learning Group, Paramus, NJ, USA

Corresponding Author:
Bridget A. Taylor, Alpine Learning Group, 777 Paramus Road, Paramus, NJ 07652, USA
Email: [email protected]
342 Behavior Modification 36(3)

discriminations (Green, 2001). Despite the importance of learning by

observation, and the significant deficits displayed by children with autism,
little applied research has been directed toward investigating strategies to
improve the observational learning skills of these children. This article will
review definitions of observational learning, the research base on teaching
responses related to observational learning, and the implications of teach-
ing these skills to children with autism who are included in typical learning
environments. We will also outline practical strategies for teachers for
improving the observational learning skills of children with autism.

Observational Learning
Observational learning is the acquisition of novel operants as a result of
observing contingencies related to the actions of others (Catania, 1998).
Observational learning was first put forth by Bandura (1977) who proposed
two types of observational learning: (a) imitation, the reproduction of a
modeled behavior, and (b) vicarious learning, the increase or decrease of an
observer’s behavior that is similar to a model based on whether the model’s
behavior is reinforced or punished. A distinction between imitation and
vicarious learning is the demonstration of the modeled behavior long after
the observer is exposed to the modeled response and its subsequent conse-
quences. Furthermore, the observer does not necessarily have to contact the
contingencies to reproduce the model. According to Bandura’s (1977) pro-
posed social learning theory, demonstration of the behavior after such a
delay in time without direct contact with contingencies can be explained by
cognitive mediation. That is, the observer forms a cognitive representation
of the observed behavior and its consequence, which then influences the
observer’s behavior at a later point in time.
Behavior analysts have proposed different explanations for observational
learning that do not imply hypothetical constructs such as those proposed by
Bandura (1977); for example, the formation of cognitive representations.
Rather, behavior-analytic explanations propose environmental variables that
can be experimentally manipulated such as the individual’s learning history
related to reinforcement for imitative behavior, the stimuli that may control
responding, stimulus generalization, and the demonstration of complex and
conditional discriminations (Deguchi, 1984; Greer, Dudek-Singer, &
Gautreaux, 2006; Masia & Chase, 1997).
According to Masia and Chase (1997), in an observational learning con-
text, the observer attends to a complex stimulus that includes a modeled
Taylor and DeQuinzio 343

response and the subsequent consequence. The modeled response and its
consequences may serve as a complex discriminative stimulus for the dem-
onstration of that response by the observer later in time. In addition, the
likelihood that a response will be imitated may be related to the observer’s
history of being reinforced for imitation and the development of a general-
ized imitative repertoire. A generalized imitative response class is necessary
to ensure the observer can imitate responses performed by others without the
need to be directly or immediately reinforced for every instance of imitation
(Kymissis & Poulson, 1994; Poulson, 2003).
Although imitation is essential for observational learning, not all imita-
tive behavior is advantageous, especially when the outcome for such
responding does not result in reinforcement. For example, we would not
want a child with autism to imitate a peer yelling out in class especially if the
peer’s teacher sends her to the principal’s office when she does so. Thus,
direct imitation does not mean that the person has learned anything about
consequences. Observational learning on the other hand implies that the
observer has learned about the consequences applied to the observed behav-
ior. This requires that the observer discriminate reinforcing and punishing
consequences, and the impact these consequences have on the observed
behavior (Greer et al., 2006). Discrimination is demonstrated by the observer
either engaging in the modeled response that was reinforced or not engaging
in the modeled response that was extinguished or punished. This type of
discrimination might be considered a conditional discrimination in that rein-
forcement for engaging in the modeled response is dependent on the pres-
ence or absence of additional stimuli (e.g., reinforcement or punishment;
Green, 2001; Masia & Chase, 1997).
In addition to discrimination, observational learning necessitates stimulus
generalization. Stimulus generalization is the spread of effects of reinforce-
ment for responses emitted in the presence of one stimulus to different but
physically similar stimuli that are not predictive of reinforcement (Stokes &
Baer, 1977). With respect to observational learning, when we observe some-
one else perform responses and those responses are followed by reinforcing
consequences, we are more likely to perform those same responses in the
presence of stimuli that are physically similar to those observed, without ever
having contacted the pleasurable consequences provided to the model (Masia
& Chase, 1997). For example, we may observe someone opening a box that
reveals candy. The modeled behavior is opening the box and the related con-
sequence is finding candy. Because we might have our own direct experience
with opening different types of boxes and finding preferred items, both which
are structurally similar to the one we observed someone else open, we might
344 Behavior Modification 36(3)

engage in this same response following observation of the model, even

though no history of reinforcement exists for opening that specific box.
Although the precise behavioral mechanisms of observational learning
have yet to be empirically validated, it is clear many different skills must
come together for observational learning to take place. To learn by observa-
tion, a child will need to attend to and observe the modeler, make complex
discriminations of another person’s actions and their outcomes, and after
a delay in time, match some properties of the modeled behavior (or not).
Thus, attending, delayed imitation, and the discrimination of contingencies
are specific skills that seem to be required for observational learning to
occur. For children with autism, these skills are often deficient
or delayed, which will invariably impede learning in environments that rely
heavily on observational learning.

Importance of Observational Learning and

Children With Autism
Most children with autism who are receiving evidence-based instruction
begin education in intensive one-to-one instruction with an adult, in a highly
structured format with well-planned and repeated learning opportunities. The
benefits of this type of instruction for children with autism have been well
documented (e.g., Hayward, Gale, & Eikeseth, 2009; Lovaas & Smith,
2003). Unfortunately, these instructional conditions may not be economi-
cally feasible for the long haul and are rarely available in general education
classrooms where group instruction is the norm. Furthermore, some children
with autism who make considerable progress in early intensive behavioral
intervention may master prerequisite skills enabling them to be included in
general education settings. Thus, it is essential that teachers and clinicians
identify strategies to shift instruction from a one-to-one context to instruction
presented in a group where learning opportunities may be diminished and
direct reinforcement for responding infrequent. This will require that the
child with autism be taught the necessary skills to learn by observing the
responses of his peers.
Observational learning also has significant social implications for children
with autism. Many “social norms” that we adhere to and follow are based on
the observed responses of others. For example, when entering a new social
situation in which we are unsure of what to do, we look around, observe what
everyone else is doing, and imitate the behavior of others to “fit in.”
Essentially, we learn the expected “norm” by observing others and the social
consequences provided to such behavior. For children with autism who
Taylor and DeQuinzio 345

present with significant social challenges, being able to learn social responses
by observing others would undoubtedly be a useful skill to enhance social
inclusion opportunities.
If we are to enhance the learning and social opportunities of children
with autism, it requires attention to remediating the skill deficits associated
with observational learning and identifying efficient strategies to enhance
observational learning skills of these children. Children with autism must,
for example, be taught to attend to others, to imitate actions after a delay,
and to identify and discriminate contingencies. Although there are only a
limited number of studies examining these specific skills in relationship to
observational learning, there is a body of literature that examines several of
these deficits in isolation and suggests ways to remediate these deficits in
children with autism.

Attending and Children With Autism

Deficits in attending such as poor or inconsistent eye contact (American
Psychiatric Association, 2000), an inability to follow eye gaze (Leekam,
Hunnisett, & Moore, 1998), not orienting to toys or materials (Donnelly,
Luyben, & Zan, 2009), and failure to engage in joint attention (Mundy &
Crowson, 1997) are understood to be some of the core diagnostic indicators
for autism. In addition, a recent investigation by Shic, Bradshaw, Klin,
Scassellati, and Chawarska (2011) used eye-tracking technology to deter-
mine that toddlers with autism attended more to background stimuli and
less to the movement of an adult and child engaged in a play interaction,
than their typically developing peers. These challenges in attending
unquestionably impact a child’s ability to learn through observation. To
learn by observing others, the child with autism will need to attend to mul-
tiple stimuli simultaneously, need to shift attention between various stim-
uli, and be required to sustain attention for long periods of time. For
example, a child with autism observing a lesson presented by a teacher
about the solar system has to attend not only to the teacher’s presentation
of the lesson material but also to the teacher’s directive or question to other
students, the student’s response to the teacher’s question, the teacher’s
consequence to the student’s response, the response of the student to the
teacher’s consequence, and any other instructional material related to the
lesson (e.g., visual display of the solar system).
Although deficits in attending for children with autism have been well
documented, there is a paucity of research examining interventions to
improve attending related to learning during instruction (Patten & Watson,
346 Behavior Modification 36(3)

2011). Most studies have addressed specific topographies of attention such as

improving joint attention (e.g., Taylor & Hoch, 2008). A few have outlined
specific reinforcement or prompting procedures to increase eye contact. For
example, a study by Donnelly et al. (2009) documented increases in attend-
ing toward task materials of a toddler with autism by using a continuous
reinforcement schedule of edibles plus praise. In another study, Tarbox,
Ghezzi, and Wilson (2006) demonstrated that a token reinforcement system
could be used to increase eye contact during discrete trial instruction. When
tokens were removed and when the backup reinforcers were no longer
available, however, eye contact decreased suggesting that a consistent
schedule of reinforcement may be necessary to maintain eye contact.
Research on teaching joint attention (see review by White et al., 2011) and
the few studies focused on increasing eye contact and attending indicate
that systematic prompting, prompt fading, and reinforcement procedures
can be used to increase these responses. Additional research is needed,
however, to identify strategies to teach children with autism to sustain
attention for longer periods of time and to examine the extent to which bet-
ter attending impacts observational learning.

Imitation and Children With Autism

Imitation is the ability to duplicate some properties of the behavior of a
model (Catania, 1998). Deficits in imitation for children with autism were
first suggested in an article by Ritvo and Provence (1953) and are now con-
sidered diagnostic features of autism (American Psychiatric Association,
2000). Empirical reviews have isolated several variables related to imitation
and autism. For example, a review by Williams, Whiten, and Singh (2004),
involving 21 studies totaling 281 children, concluded that imitative deficits
were more apparent in younger children with autism. Furthermore, they
determined that children with autism had the most difficulty imitating non-
meaningful gestures and had the least difficulty imitating actions with
objects. A recent longitudinal study by Young, Roger, Hutman, Rozga, and
Ozonoff (2011) revealed that children with familial risk for autism and later
diagnosed with autism showed delayed imitation skills but were indistin-
guishable from other high-risk infants who showed other cognitive delays
not related to autism. The study also showed that all groups displayed simi-
lar linear increases in imitation between 12 and 24 months, and these
increases were associated with individual growth in expressive language
and social engagement ratings. In another study by Young, Krantz,
Taylor and DeQuinzio 347

McClannahan, and Poulson (1994), it was determined that learning a gener-

alized imitative repertoire for one type of imitative response (e.g., vocal) did
not necessarily result in the emergence of another type of imitative response
(e.g., motor), suggesting that various topographies of imitation may be dis-
tinct response classes. Collectively, these studies confirm that children with
autism have delays or deficits in imitation and that these impairments may
be related to certain topographies of imitation and other skills, such as
expressive language and social behavior. Furthermore, imitation seems to
improve with age and may show a sequential linear progression.
Despite deficits in imitation, it has been demonstrated that children with
autism can be taught to imitate a variety of responses. A number of methods
to teach imitation have been investigated such as discrete trial teaching
(Lovaas, 1977), the use of reciprocal imitation in play contexts (Ingersoll &
Schreibman, 2006), and prompting and reinforcing imitation in a group
(Garfinkle & Schwartz, 2002). Early research documented the benefits of
prompting and differential reinforcement to teach imitation of motor and
vocal behavior demonstrated by an adult (e.g., Baer, Peterson, & Sherman,
1967; Metz, 1965). Subsequent studies have documented that children with
autism can learn to imitate an array of complex responses, including facial
expressions (DeQuinzio, Townsend, Sturmey, & Poulson, 2007), gestures
(Buffington, Krantz, McClannahan, & Poulson, 1998; Ingersoll, Lewis, &
Kroman, 2007), play responses of peers (Garfinkle & Schwartz, 2002) and
actions, and vocalizations modeled on videos (D’Ateno, Mangiapanello, &
Taylor, 2003; Taylor, Levin, & Jasper, 1999).
Delayed imitation, reproducing an observed action after a period of time
has elapsed (Garcia, 1976), is an essential skill related to observational
learning. To learn by observation, a person will need to be able to observe
a response demonstrated by a model and be able to produce that response
later in time in the presence of similar stimuli. Researchers have found that
typically developing 6-month-olds can imitate after a 24-hr delay, 12-month-
olds after a 4-week delay, and 2-year-olds after a 4-month delay (Barr,
Dowden, & Hayne, 1996). To date, there are no systematic studies examin-
ing delayed imitation in children with autism. But some researchers have
documented that some children with autism can produce responses after
observing a model following a period of time (Hobson & Lee, 1999).
Nonetheless, additional research on teaching delayed imitation is necessary
to guide practice on how best to facilitate these responses in relationship to
observational learning in children with autism.
348 Behavior Modification 36(3)

Discriminating Contingencies and Children

With Autism
As noted above, behavior analysts have argued that observational learning
requires complex discriminations, often characterized as conditional dis-
criminations (Masia & Chase, 1997). One such discrimination necessary for
observational learning is identifying and responding accordingly to a rein-
forcing and a punishing consequence delivered to another. Making these
discriminations can be challenging for a child with autism who may have
difficulty responding to vocal verbal behavior related to positive conse-
quences (e.g., someone saying, “Good job!”) and punishing consequences
(e.g., someone saying, “No, that’s incorrect.”), and in responding to nonvo-
cal responses such as displays of emotion (e.g., someone grimacing or
frowning). For example, if a child observes a peer playing with a door by
swinging it open and closed, and the peer accidently gets his finger caught
in the door and grimaces and whines, the observing child has to infer, based
on the peer’s reaction (i.e., facial grimacing and whining), that the conse-
quence of playing with the door (i.e., getting his finger caught) is unpleasant
and as a result he should not play with a door by swinging it open and
closed. This inference will no doubt be difficult for the child with autism
who has deficits in discriminating and responding to emotional reactions of
others (Baron-Cohen, Leslie, & Frith, 1985; Weeks & Hobson, 1987).
The challenges that children with autism have with respect to learning
conditional discriminations are reflected by a vast body of research dedicated
to evaluating the best strategies for teaching these repertoires (Fisher, Kodak,
& Moore, 2007; Green, 2001; Pérez-González & Williams, 2002; Williams,
Pérez-González, & Queiroz, 2005). Although prior research has demon-
strated that children with autism can learn to discriminate both between and
among classes of stimuli such as colors, people, and common objects, few
studies have attempted to apply these strategies to teach children with autism
to respond to the complex stimulus relations involved in observational
learning.
Of notable exception is a study conducted by Pereira-Delgado and Greer
(2009) in which two 5-year-old children with autism were taught to discrimi-
nate between correct and incorrect responses of a peer by monitoring both the
responses of the peer and the consequences delivered by the teacher.
Specifically, participants with autism were taught to choose a green block
when a peer‘s correct response was reinforced by the teacher and a red block
when a peer‘s response was corrected by the teacher. The researchers argued
that learning to discriminate the responses of the peer and the consequences
Taylor and DeQuinzio 349

related to those responses facilitated the acquisition of novel sight words.

This study holds promise that children with autism may be able to learn to
discriminate these consequence and as a result learn novel responses. More
research is needed, however, to examine the effects of learning these dis-
criminations related to more complex responses and to clarify best practice
procedures to enhance these skills in children with autism.

Studies Evaluating Observational Learning and

Children With Autism
Most research on observational learning has been conducted with typically
developing children or children with general developmental disabilities. Few
have included children with autism. Observational learning with children
with other disabilities has been assessed to teach a variety of skills, including
reading of sight words (Schoen & Ogden, 1995), completing long response
chains (Werts, Caldwell, & Wolery, 1996), seeking assistance following
injury (Christensen, Lignugaris-Kraft, & Fiechtl, 1996), preparing food
(Griffin, 1992), and completing language tasks (Goldstein & Mousetis,
1989). Research on observational learning, specifically with children with
autism, reveals inconsistent findings both within and across studies. Some
children with autism acquired new responses by observing a model, whereas
others did not. In addition, most studies examined the effects of the child
with autism observing a competent model perform the responses accurately
as opposed to watching the model learn a new response. In addition, only a
few have addressed teaching components of observational learning such as
attention to the modeled response (Taylor, DeQuinzio, & Stine, in review)
and teaching discrimination of contingencies (Pereira-Delgado & Greer,
2009). Nonetheless, existing studies offer some support that some children
can learn by observing responses demonstrated by others.
One of the first studies examining observational learning and children
with autism was a study by Varni, Lovaas, Koegel, and Everett (1979). In this
study, children with autism observed an adult model perform responses and
receive reinforcement by another adult for performing the responses accu-
rately. It was found that children with autism performed similarly to the
youngest age group. This was different for the children without autism whose
observational learning progressed with the children’s chronological age. For
the participants with autism in this study, chronological age did not determine
how much information they acquired as a result of observing a model. Rather,
responding was idiosyncratic. The authors concluded that failure to learn by
350 Behavior Modification 36(3)

observation may be related to stimulus over selectivity, and as a result, the

children with autism were unable to attend to the relevant features of the task.
In another study, Egel, Richman, and Koegel (1981) found performance
on five discrimination tasks improved for four children with autism as a
result of watching typically developing peers model the responses. In this
study, the children observed the peer models perform the responses accu-
rately and receive reinforcement for doing do so. As a result, accuracy of
the discrimination task improved for all four children with autism. Tryon
and Keane (1986) found that “autistic like” children learned to engage
appropriately with a novel toy by observing a peer model appropriate play
with the toy. Similarly, Charlop, Schreibman, and Tryon (1983) found that
four children with autism learned new receptive language tasks as a result
of observing competent learners with autism demonstrate those responses
with an adult. A study by Kamps, Walker, Locke, Delquardia, and Hall
(1990) on the other hand showed varied rates of acquisition of a sight word
reading task presented in a group observational learning context.
A more recent study by Rehfeldt, Latimore, and Stromer (2003) examined
the formation of stimulus classes of reading skills, in three individuals with
developmental disabilities (one of whom had autism), as a result of observing
a model demonstrate conditional discriminations. They found that all partici-
pants subsequently demonstrated full stimulus classes with the stimuli
involved in direct training; on the other hand, based on the observed model,
none of the participants demonstrated full stimulus classes. In a second
experiment, however, participants demonstrated full stimulus classes with at
least one of the model’s set of training stimuli.
Taylor, DeQuinzio, and Stine (in review), investigated teaching three chil-
dren with autism to monitor the responses of a peer model during a site word
reading task. The monitoring response consisted of imitating the peer’s
response when asked by the teacher, “What did (peer’s name) say?” (after the
peer read the word), and matching a plastic chip to the word that was read on
a word-matching grid to indicate the child was looking at the instructional
stimuli. In one condition, the monitoring response was required and in another
other condition, with a different set of words, the monitoring response was not
required. Separate tests of the two sets of words revealed that accuracy for
reading was better for the words where the monitoring response was required.
Eventually, for two participants, accuracy for the words in the nonmonitoring
condition also increased, indicating that there may have been generalization of
the monitoring response to the conditions where monitoring stimuli were not
present. For the third participant, accuracy of reading the site words did not
improve until the monitoring response was required, indicating that for this
Taylor and DeQuinzio 351

participant, consistent prompts of visually monitoring and imitating the

responses of the peer were necessary to learn the new words. This study sup-
ports the notion that children with autism may not acquire novel information
by simply being exposed to models performing responses and receiving rein-
forcement for those responses. Instead, children with autism may require
prompting to visually attend to instructional stimuli and to imitate the response
of their peers.
Collectively, these studies indicate that some children with autism will
learn new responses as a result of observing proficient models perform
those responses. Unknown, however, is why some children with autism are
able to do so and why some are not. It stands to reason that competent
observational learners likely demonstrate specific prerequisite skills that
facilitate learning through observation. It may be the case, for example, that
the skilled observational learners attend to models and instructional stimuli
for appropriate durations, engage in delayed imitation, and can discriminate
contingencies. Additional research is needed to identify the conditions
under which children with autism learn through observation and the essen-
tial skills required to do so.

Strategies for Increasing Observational

Learning
As reviewed above, there are a number of potential skills which when defi-
cient may preclude learning through observation. Although more research is
needed to demonstrate the particular prerequisite skills and the most efficient
strategies to teach them, we offer practical strategies that may increase
attending, imitation, and the discrimination of contingencies. Some of the
strategies we offer are based on empirical research and others on clinical
practice. Although a benefit of observational learning is to increase new
responses without the direct instruction of each individual learner, building
the prerequisite responses will initially require individualized instruction.
General classroom strategies can also be used to facilitate these skills in a
group instruction format.
Teach sustained attention to peer models. The child with autism must first
demonstrate proficiency in looking at, or orienting toward, a model for an
extended duration so that he may observe the entire response performed by
the model and subsequent consequences. To date, there is no research specifi-
cally evaluating strategies for increasing sustained attention by children with
autism. Shaping, a procedure in which successive approximations of a target
response are differentially reinforced (Cooper, Heron, & Heward, 2007), could
352 Behavior Modification 36(3)

be an effective strategy for increasing sustained attention. Teachers, at first,

may reinforce short instances of looking at, or orienting toward, a model and
then systematically increase the duration required for reinforcement. Initially,
to shape attention toward a model, the teacher could have the child with
autism sit across from a peer and provide an instruction to the child to look at
the peer model (e.g., the teacher could say, “Look at John”) and use a gesture
prompt of pointing in the direction of the peer. When the child looks in the
direction of the peer, the teacher would provide praise and a preferred snack
or toy to reinforce his response of looking toward the peer. To ensure the
child is actually looking at the peer, the teacher could ask the child questions
about the peer’s responses or behavior (e.g., ask, “What is John doing?”). The
child’s correct answer ensures that he has looked at the peer. To increase the
likelihood that the child will look at the peer, the teacher could have the peer
model fun or interesting actions or activities (such as demonstrating play with
a fun toy). Over time, to shape longer durations of looking at the peer model,
the teacher could have the peer model perform a sequence of actions out of
the chair and with other peers or teachers. The teacher could use a timer to
cue the expected duration of attending, which can be systematically increased
over time. As the child is able to sustain attention to the peer for longer peri-
ods of time, the teacher could add distractor stimuli, such as other students or
toys, that may divert the child’s attention and teach the child to sustain atten-
tion to the model even when these distracting stimuli are present.
Promote generalized imitation of peer’s vocal and motor responses. Imitation
is when the child’s behavior is contiguous to the behavior of a model and is
topographically similar to that of the model (Baer et al., 1967). As described
earlier, research has illustrated that prompting and reinforcement procedures
can be used to shape imitative repertoires in children with autism using adult
models. Arguably, peer imitation is an essential prerequisite skill for observa-
tional learning. Studies published using peers as models are less abundant;
however, they do provide a framework for developing peer imitation training
procedures (Carr & Darcy, 1990; Ganz, Bourgeois, Flores, & Campos, 2008;
Garfinkle & Schwartz, 2002). To teach imitation of peer responses, a teacher
may initially have the child sit across from a peer at a table or in a play area.
The teacher would then ask the peer to demonstrate actions (e.g., the peer is
instructed to push a car back and forth) and would present an instruction to
direct the child to imitate the peer (e.g., the teacher would say, “Do what John
is doing”). If necessary, the teacher could guide the child to imitate the peer’s
action. When the child does, the teacher would provide praise and a preferred
snack or toy, to reinforce the imitative response. To promote generalization,
the teacher would have the peer model different actions each time, until the
Taylor and DeQuinzio 353

child can imitate novel actions without any prompting or reinforcement. To

increase the likelihood that the child with autism will imitate the peer, the
teacher could have the peer model actions associated with preferred toys that
the child may not know how to operate. Or the teacher can have the peer
model actions that lead to a desired outcome for the child with autism. For
example, the teacher may present a new toy to the child and when the child
shows interest in the toy, but is unable to operate it, the teacher could have the
child observe the peer manipulate the toy to know how it works. The teacher
would then provide the child with autism an opportunity to imitate the action
with the toy.
In addition to this structured imitation training with peers, Brown, Brown,
and Poulson (2008) argued that it is also important to teach children with
autism to imitate the responses of peers without a verbal instruction. The
authors reasoned that in more “ordinary” learning environments, such as
general education classrooms, verbal instructions to imitate peers and pro-
grammed consequences for doing so are not likely to occur. Thus, structured
imitation training should attempt to establish imitation of peers under condi-
tions that are similar to those involved in observational learning. This can be
promoted by eventually teaching imitation of peers in a variety of contexts
and eventually fading or no longer using directives for imitation, such as
“Do what John is doing.”
Finally, the child with autism will have to demonstrate a response mod-
eled by a peer following some delay (Garcia, 1976). To teach this, the teacher
could present novel actions for the child to imitate and impose a delay of
time between the modeled action and the opportunity for the child to display
the response. For example, the teacher could have the child observe the peer
demonstrating a specific action with a novel toy and wait 15 minutes before
giving the toy to the child with autism to assess whether the child demon-
strates the responses modeled by the peer. This same sequence can be con-
ducted to teach the child with autism to imitate vocal behavior. For example,
the child could observe a peer answering a teacher’s question (e.g., the peer
says, “four” in response to the teacher’s question, “How much is two plus
two?”), and the teacher could wait a few minutes and then present the ques-
tion to the child with autism to see whether he imitates what was modeled by
the peer (i.e., says, “four”).
Teach discrimination of consequences. Finally, and arguably, the most com-
plex component response of observational learning is discriminating the
consequences of the responses of others. This requires that the learner can
respond differentially to complex stimuli, that is, the modeled response
(e.g., a correct answer) as well as the consequences associated with each
354 Behavior Modification 36(3)

modeled response (e.g., teacher praise). Within the observational learning

paradigm, learners must match the responses of the model that were rein-
forced and refrain from engaging in responses that were not. Because of the
complexity of the discrimination, it might be helpful to first teach learners to
discriminate reinforced from nonreinforced responses by engaging in an
arbitrary response (e.g., pointing to a red card versus pointing to a green
card; see Pereira-Delgado & Greer, 2009). Requiring a simple response,
such as pointing, initially will help teachers determine whether the learner
with autism can discriminate among the complex stimuli without requiring
additional verbal behavior. For example, initially, the teacher could seat the
child in view of a peer modeling both correct and incorrect responses to a
lesson. When the model engages in a correct response as indicated by teacher
praise (e.g., “Excellent, you are correct!”), the teacher could prompt the
child to point to the green card. When the model engages in an incorrect
response, as indicated by the teacher’s corrective statement (e.g., “No, that’s
not correct”), the teacher would not provide praise and would prompt the
child to point to the red card. Eventually, the teacher would remove all
prompts until the child accurately points to the green card when the model is
correct and the red card when the model is incorrect.
Have the child practice the skills to learn new information. Following mastery
of the above prerequisites, teachers may consider implementing the following
observational learning sequence: The teacher would first identify a skill the
child cannot demonstrate but a peer can (e.g., reading of sight words, new
vocabulary, answers to general knowledge questions, etc.). The child would be
prompted to observe a peer engaged in an instructional session with a teacher
related to the novel response (e.g., the teacher asks the peer to read sight words).
The child should be able to observe the instructional stimuli that the peer sees
(e.g., the child should see the sight words). After each correct response demon-
strated by the model, the teacher could turn to the child with autism and ask him
to repeat what the peer just said or to imitate what the peer just did. When the
child responds correctly, the teacher could provide praise and preferred stimuli
(e.g., toys, stickers, tokens). Later (e.g., 15 min to a half-hour later), the teacher
could test the child on the responses modeled by the peer (e.g., the teacher asks
the child to read the sight words) to determine whether he is learning the new
responses as a result of observing the peer demonstrate the responses (Taylor et
al., 2012). Assessing the response in the absence of the model tests the extent
to which the observer has acquired novel responses as a result of observing the
model.
Implement general classroom strategies. In addition to teaching the skills
individually, teachers can promote the skills during group lessons throughout
Taylor and DeQuinzio 355

the school day (Taylor, 2013). For example, to increase attending of the child
with autism toward peers, during a group lesson, the teacher could monitor
the child’s attention and provide directives to the child to look at peers per-
forming actions. The teacher could present the instructions directly to the
child with autism (e.g., by saying, “Billy, look at what Peter is doing”) or to
the entire group (e.g., by saying, “Everyone look up here”). In addition, to
ensure the child with autism is attending to the peer, the teacher might ask the
child to recall or name actions performed by the peer immediately after the
peer performs the response. For example, if a peer is called to the front of the
room to complete a math problem, the teacher could ask the child with autism
to say what the peer did (e.g., by saying, “Billy, what problem did Peter just
complete?”).
To encourage imitation of peer responses, if the child is unable to answer
a question correctly, the teacher should call on a peer to model the correct
answer, reinforce the peer, and then call on the child again to see whether he
imitates the correct response of the peer. In addition, to encourage the child
with autism to attend to the consequences provided to the peer, teachers
should be explicit when providing consequences to the peer’s response (e.g.,
when the peer is correct, say, “You are right! The capital of New York is
Albany” and when the peer is incorrect, say, “No that’s not correct. The
Capital of New York is not Trenton”). The teacher can then assess whether
the child with autism is discriminating the consequences by asking the child
if the peer was correct or not. To promote generalization, the teacher could
implement these strategies with a variety of lesson types and instructional
stimuli.
Learning by observing others is an essential skill, but one that may not
come so easily for children with autism. Behavior-analytic explanations of
observational learning provide a framework for experimental evaluation of
the mechanisms associated with observational learning (Deguchi, 1984;
Masia & Chase, 1997) and for the remediation of observational learning
deficits in individuals with severe developmental disabilities, such as autism
(Greer et al., 2006; Taylor et al., 2012). If we are to increase the learning
opportunities of children with autism in typical learning environments, clini-
cal practice must incorporate learning objectives and instructional strategies
to develop the skills necessary to learn by observation. More research is
certainly necessary to identify all of the responses required for observational
learning as well as efficient methodologies. Nevertheless, an emerging body
of literature is available that outlines strategies to improve the observational
learning skills of children with autism. By improving such skills, we will
undoubtedly enhance the educational and social opportunities for children
with autism.
356 Behavior Modification 36(3)

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or
publication of this article.

References
American Psychiatric Association. (2000). Diagnostic and statistical manual of
mental disorders (4th ed., text rev.). Washington, DC: Author.
Baer, D., Peterson, R. F. M., & Sherman, J. A. (1967).The development of general-
ized imitation by reinforcing behavioral similarity to a model. Journal of the
Experimental Analysis of Behavior, 10, 405-416.
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice Hall.
Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a
“theory of mind”? Cognition, 21, 37-46.
Barr, R., Dowden, A., & Hayne, H. (1996). Developmental changes in deferred
imitation by 6-to 24-month-old infants. Infant Behavior & Development, 19,
159-171.
Brown, A. K., Brown, J. L., & Poulson, C. L. (2008). Discriminating which fork
to use: Teaching selective imitation to people with autism. Research in Autism
Spectrum Disorders, 2, 199-208. doi:10.1016/j.rasd.2007.06.001
Buffington, D. M., Krantz, P. J., McClannahan, L. E., & Poulson, C. L. (1998).
Procedures for teaching appropriate gestural communication skills to children
with autism. Journal of Autism and Developmental Disorders, 28, 535-545.
doi:10.1023/A:1026056229214
Carr, E. G., & Darcy, M. (1990). Setting generality of peer modeling in children with
autism. Journal of Autism and Developmental Disorders, 20, 45-59. doi:10.1007/
BF02206856
Catania, A. C. (1998). Learning (interim 4th ed.). Cornwall-on-Hudson, NY: Sloan.
Charlop, M. H., Schreibman, L., & Tryon, A. S. (1983). Learning through observation:
The effects of peer modeling on acquisition and generalization in autistic children.
Journal of Abnormal Child Psychology, 11, 355-366. doi:10.1007/BF00914244
Christensen, A., Lignugaris-Kraft, B., & Fiechtl, B. J. (1996). Teaching pairs of
preschoolers with disabilities to seek adult assistance in response to simulated
injuries: Acquisition and promotion of observational learning. Education &
Treatment of Children, 19, 3-18.
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis
(2nd ed.). Upper Saddle River, NJ: Pearson.
Taylor and DeQuinzio 357

D’Ateno, P., Mangiapanello, K., & Taylor, B. A. (2003). Using video modeling to
teach complex play sequences to a preschooler with autism. Journal of Positive
Behavior Interventions, 5, 5-11.
Deguchi, H. (1984). Observational learning from a radical-behavioristic viewpoint.
Behavior Analyst, 7, 83-95.
DeQuinzio, J., Townsend, D., Sturmey, P., & Poulson, C. L. (2007). Generalized
imitation of facial models by children with autism. Journal of Applied Behavior
Analysis, 40, 755-759.
Donnelly, J. I., Luyben, P. D., & Zan, C. S. (2009). Increasing eye contact toward
learning materials in a toddler with autism. Journal of Prevention & Intervention
in the Community, 37, 170-176.
Egel, A. L., Richman, G., & Koegel, R. L. (1981). Normal peer models and autistic
children’s learning. Journal of Applied Behavior Analysis, 14, 3-12.
Fisher, W. W., Kodak, T., & Moore, J. W. (2007). Embedding an identity-matching
task within a prompting hierarchy to facilitate acquisition of conditional dis-
criminations in children with autism. Journal of Applied Behavior Analysis, 40,
489-499.
Ganz, J. B., Bourgeois, B. C., Flores, M. M., & Campos, B. (2008). Implementing
visually cued imitation training with children with autism spectrum disorders
and developmental delays. Journal of Positive Behavior Interventions, 10, 56-66.
doi:10.1177/1098300707311388
Garcia, E. E. (1976). The development and generalization of delayed imitation.
Journal of Applied Behavior Analysis, 9, 499.
Garfinkle, A. N., & Schwartz, I. S. (2002). Peer imitation: Increasing social inter-
action in children with autism and other developmental disabilities in inclusive
preschool classrooms. Topics in Early Childhood Special Education, 22, 26-38.
Goldstein, H., & Mousetis, L. (1989). Generalized language learning by children
with severe mental retardation: Effects of peers’ expressive modeling. Journal of
Applied Behavior Analysis, 22, 245-259.
Green, G. (2001). Behavior analytic instruction for learners with autism: Advances
in stimulus control technology. Focus on Autism and Other Developmental
Disabilities, 16, 72-85.
Greer, R., Dudek-Singer, J., & Gautreaux, G. (2006). Observational learning. Inter-
national Journal of Psychology, 41, 486-499. doi:10.1080/00207590500492435
Griffin, A. K. (1992). Triadic instruction of chained food preparation response:
Acquisition and observational learning. Journal of Applied Behavior Analysis,
25, 193-204.
Hayward, D. W., Gale, C. M., & Eikeseth, S. (2009). Intensive behavioural interven-
tion for young children with autism: A research-based service model. Research
in Autism Spectrum Disorders, 3, 571-580.
358 Behavior Modification 36(3)

Hobson, R., & Lee, A. (1999). Imitation and identification in autism. Journal of Child
Psychology and Psychiatry, 40, 649-659.
Ingersoll, B., Lewis, E., & Kroman, E. (2007). Teaching the imitation and spontane-
ous use of descriptive gestures to young children with autism using a naturalistic
behavioral intervention. Journal of Autism and Developmental Disorders, 37,
1446-1456.
Ingersoll, B., & Schreibman, L. (2006). Teaching reciprocal imitation skills to
young children with autism using a naturalistic behavioral approach: Effects on
language, pretend play, and joint attention. Journal of Autism and Developmental
Disorders, 36, 487-505.
Kamps, D., Walker, D., Locke, P., Delquardia, J., & Hall, R. V. (1990). A comparison
of instructional arrangement for children with autism served in a public school
setting. Education & Treatment of Children, 13, 197-216.
Kymissis, E., & Poulson, C. L. (1994). Generalized imitation in preschool boys.
Journal of Experimental Child Psychology, 58, 389-404.
Ledford, J. R., Gast, D. L., Luscre, D., & Ayres, K. M. (2008). Observational and
incidental learning by children with autism during small group instruction.
Journal of Autism and Developmental Disorders, 38, 86-103. doi:10.1007/
s10803-007-0363-7
Leekam, S. R., Hunnisett, E., & Moore, C. (1998). Targets and cues: Gaze-following
in children with autism. Journal of Child Psychology and Psychiatry, 39, 951-962.
Lovaas, O. (1977). The autistic child: Language development through behavior
modification. Oxford, England: Irvington.
Lovaas, O., & Smith, T. (2003). Early and intensive behavioral intervention in autism.
In A. E. Kazdin, J. R. Weisz, A. E. Kazdin, & J. R. Weisz (Eds.), Evidence-based
psychotherapies for children and adolescents (pp. 325-340). New York, NY:
Guilford.
Masia, C. L., & Chase, P. (1997). Vicarious learning revisited: A contemporary
behavior analytic interpretation. Journal of Behavior Therapy and Experimental
Psychiatry, 28, 41-51.
Metz, J. R. (1965). Conditioning generalized imitation in autistic children. Journal of
Experimental Child Psychology, 2, 389-399.
Mundy, P., & Crowson, M. (1997). Joint attention and early social communication:
Implication for research on intervention with autism. Journal of Autism and
Developmental Disorders, 27, 653-676.
Patten, E., & Watson, L. R. (2011). Interventions targeting attention in young children
with autism. Journal of Speech-Language Pathology, 20, 60-69.
Pereira-Delgado, J. A., & Greer, R. D. (2009). The effects of peer monitoring training
on the emergence of the capability to learn by observing instruction received by
peers. Psychological Record, 59, 407-434.
Taylor and DeQuinzio 359

Pérez-González, L., & Williams, G. (2002). Multicomponent procedure to teach

conditional discriminations to children with autism. American Journal on Mental
Retardation, 107, 293-301.
Poulson, C. L. (2003). Generalized imitation: An empirical and conceptual analysis.
In K. S. Budd & T. Stokes (Eds.), A small matter of proof: The legacy of Donald
M. Baer. Reno, NV: Context Press.
Rehfeldt, R. A., Latimore, D., & Stromer, R. (2003). Observational learning and the
formation of classes of reading skills by individuals with autism and other devel-
opmental disabilities. Research in Developmental Disabilities, 24, 333-358.
Ritvo, S., & Provence, S. (1953). Form perception and imitation in some autistic
children. Psychoanalytic Study of the Child, 8, 155-161.
Schoen, S. F., & Ogden, S. (1995). Impact of time delay, observational learning, and
attention cuing upon word recognition during integrated small-group instruction.
Journal of Autism and Developmental Disorders, 25, 503-519.
Senju, A., Yaguchi, K., Tojo, Y., & Hasegawa, T. (2003). Eye contact does not facili-
tate detection in children with autism. Cognition, 89, B43-B51.
Shic, F., Bradshaw, J., Klin, A., Scassellati, B., & Chawarska, K. (2011). Limited
activity monitoring in toddlers with autism spectrum disorder. Brain Research,
1380, 246–254. doi:10.1016/j.brainres.2010.11.074
Smith, I. M., & Bryson, S. E. (1994). Imitation and action in autism: A critical review.
Psychological Bulletin, 116, 259-273.
Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal
of Applied Behavior Analysis, 10, 349-367.
Tarbox, R. S. F., Ghezzi, P. M., & Wilson, G. (2006). The effects of token reinforce-
ment on attending in a young child with autism. Behavioral Interventions, 21,
155-164.
Taylor, B. A. (2013). “Do this, but don’t do that”: Teaching children with autism
to learn by observation. In W. L. Heward (Ed.), Exceptional children: An intro-
duction to special education (10th ed., pp. 240-242). Upper Saddle River, NJ:
Pearson.
Taylor, B. A., DeQuinzio, J. A., & Stine, J. (in review) Increasing observational learn-
ing by children with autism: A preliminary analysis.
Taylor, B. A., & Hoch, H. (2008). Teaching children with autism to respond to
and initiate bids for joint attention. Journal of Applied Behavior Analysis, 41,
377-391.
Taylor, B. A., Levin, L., & Jasper, S. (1999). Increasing play-related statements in
children with autism toward their siblings: Effects of video modeling. Journal of
Developmental and Physical Disabilities, 11, 253-264.
360 Behavior Modification 36(3)

Tryon, A. S., & Keane, S. P. (1986). Promoting imitative play through generalized
observational learning in autistic-like children. Journal of Abnormal Child Psy-
chology, 14, 532-549.
Varni, J. W., Lovaas, O. I., Koegel, R. L., & Everett, N. C. (1979). An analysis of
observational learning in autistic and normal children. Journal of Abnormal Child
Psychology, 7, 31-43.
Weeks, S. J., & Hobson, R. P. (1987). The salience of facial expressions for autistic
children. Journal of Child Psychology and Psychiatry and Allied Disciplines, 28,
137-152.
Werts, M. G., Caldwell, N. K., & Wolery, M. (1996). Peer modeling of response
chains. Journal of Applied Behavior Analysis, 29, 53-66.
White, P. J., O’Reilly, M., Streusand, W., Levine, A., Sigafoos, J., Lancioni, G.,
. . . Aguilar, J. (2011). Best practices for teaching joint attention: A systematic
review of the intervention literature. Research in Autism Spectrum Disorder, 5,
1283-1295.
Williams, G., Pérez-González, L., & Queiroz, A. (2005). Using a combined blocking
procedure to teach color discrimination to a child with autism. Journal of Applied
Behavior Analysis, 38, 555-558.
Williams, J. H. G., Whiten, A., & Singh, T. (2004). A systematic review of action
imitation in autistic spectrum disorder. Journal of autism and Developmental Dis-
orders, 34, 285-299.
Young, J. M., Krantz, P. J., McClannahan, L. E., & Poulson, C. L. (1994). General-
ized imitation and response-class formation in children with autism. Journal of
Applied Behavior Analysis, 27, 685-697.
Young, G. S., Roger, S. J., Hutman, T., Rozga, A., & Ozonoff, S. (2011). Imitation
from 12 to 24 months in autism and typical development: A longitudinal Rasch
analysis. Developmental Psychology, 47, 1565-15798.

Bios
Dr. Bridget A. Taylor is Co-founder and Executive Director of Alpine Learning
Group. She has specialized in the education and treatment of children with autism for
the past twenty-five years. Dr. Taylor is active in the autism research community and
has published numerous publications on effective interventions for autism.

Dr. Jaime A. DeQuinzio is a Board Certified Behavior Analyst and has specialized
in the education and treatment of children with autism for over 13 years. She has
published in the areas of generalized imitation and the amelioration of social-affective
deficits and problem behavior of children with autism.