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

net/publication/263216725

Incidental Teaching of Age-Appropriate Social Phrases to Children

With Autism

Article  in  Research and Practice for Persons with Severe Disabilities · June 2007
DOI: 10.2511/rpsd.32.2.112

CITATIONS READS
45 913

2 authors:

Gail G. Mcgee Teresa Daly

Emory University University of Central Florida
34 PUBLICATIONS   4,899 CITATIONS    12 PUBLICATIONS   430 CITATIONS   

SEE PROFILE SEE PROFILE

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

Walden View project

Improving Post-secondary Education Outcomes for Hispanic/Latino Youth with Autism Spectrum Disorders View project

All content following this page was uploaded by Gail G. Mcgee on 21 January 2021.

The user has requested enhancement of the downloaded file.

Research & Practice for Persons with Severe Disabilities copyright 2007 by
2(X)7. Vol. .12. No. 2. 112-123 TASH

Incidental Teaching of Age-Appropriate

Social Phrases to Children With Autism
Gail G. McGee and Teresa Daly
Emory University

Successful inctusion is facilitated when children with intervention time. Among the first contributions of be-
autism fit in and interact in meaningful ways with their havior analysis researchers was a sizeable empirical ht-
typical peers. However, deficits in conversational lan- erature on proeedures for teaching the bread-and-butter
guu^e likely limit the social aitractiveness of children wilh responses of functional language (Guess. Sailor, & Baer.
autism to their classmates. This study evaluated an in- 1976; Hart & Risley, 1968: Sehreibman, 1988), and the
cidenial teaching approach to promoting use of age- cumulative impact of this body of detailed intervention
appropriate social phrases by three preschool-aged hoys procedures has been to make it possible to teach nearly
with atttism. A multiple baseline design demonstrated that all of a formerly mute population how to talk. If sim-
introduction of incidental teaching yielded immediate use ilarly vigorous effort is now invested in developing a
ofthe targeted social phrases ("All right'' and "You know comprehensive portfolio of strategies for teaching social
what?") during instructional se.ssions, and children communication to ehiidren with autism, the result could
transferred use of the new social phrases to unprompted be a significant impact on the .social challenges that de-
conditions Initial instruction required somewhat artifi- fine autism.
cial stimulus-response relationships lo compemale for A foundation has been established upon which an
the lack of responsiveness to social consequences that evidence-based social communication curriculum for chil-
characierlze.s autism. However, as in other studies in dren with autism may be built. For example, proeedures
which conversational language was successfully tatight to have been developed and proven effective in teaching
children with autism, systematic fading procedures fa- children with autism how to answer wh-concept questions
cilitated transfer of .stimtilus control from the contrived while reporting on temporally remote events (Krantz,
cues needed to teach a child wilh autism lo sav age- Zaienski. Hall, Fenske, & McClannahan. 1981). Replica-
appropriate social phrases lo conditions thai call for ble procedures have also been developed for teaching
comments and queries in everyday siiuaiions. yes/no discriminations to children with autism; when taught
to answer yes or no to questions about their preferences
DESCRIPTORS: autism, social eommunication, (i.e., "Do you want a cookie?"), children with autism
language instruetion were also able to answer yes or no to questions about
The power of early intervention and the trend for chil- objeet labels (i.e., "Is this a cookie?") (Neef, Walters, &
dren with autism to enter treatment at very young ages Egei. 1984).
make it increasingly feasible to teach children with au- Social language is essentially a prerequisite for learn-
tism how to have conversations with others. However, ing how to interact with peers (Haring & Breen. 1992).
it may be argued that the field of applied behavior anal- Various components of peer-related conversational skills
ysis has not yet fully exploited the potential inherent in have been successfully taught to children with autism, in-
the availability of a longer eourse of prime early autism cluding (I) eompliments on good gamesmanship (McGee,
Krantz. & McClannahan. 1984); (2) statements of affec-
tion (Charlop & Walsh. 1986; McEvoy et al.. 1988); and
Pilot studies associated with this research were partially (3) scripted play-related responses (Odom & Strain, 1986).
supported by Grant #HO23C4O113t)96. OSERS. US. Depart-
ment of Education. The opinions expressed herein do not Progress notwithstanding, the promotion of general-
neeessarily reflect the position or policy of the U.S. Department ized conversational language in children with autism
of Education. remains a complex procedural challenge. Skinner (19.'i7)
The authors express appreciation to the Walden teaching described much of verbal behavior as a response to
staff for devising the practical applications of these procedures. environmental and social cues, which is maintained by
In addition, Lecsa Mann. PhD. and Mary Lynn Boscardin,
EdD, made substantial contributions toward initial develop- social eonsequences. One of the challenges of teaching
ment of the verbatim language data collection methods used in social conversation to children with autism is the intrin-
this study. sic requirement for flexible use of language outside the
Address al! correspondence and reprint re-quests to Gail G. confines of the stimulus conditions in whieh it was ini-
McGee, PhD. Emory Autism Center. Department of Psychi- tially taught. Another major barrier to overcome in pro-
atry and Behavioral Sciences, Emory University School of
Medicine, I.'5.'>1 Shoup Court, Clairmont Campus, Atlanta. GA moting use of social language by ehiidren with autism is
30322. E-mail: gmegee@emory.edu that social communication is normally elicited and main-
112
 Age-Appropriate Social Phrases

tained by an interest in social interaction; not only do itations of preschool jokes. In the current study, jokes
children with autism share a nearly universal lack of re- were replaced with child-culture social phrases that were
sponsiveness to social consequences, but they also have validated as pleasing to adults.
characteristic initiation difficulties. Parents of typical children take special delight in re-
Incidental teaching procedures may offer promise as lating their children's cute sayings, whereas parents of
a means of promoting use of conversational language children with autism seldom get to call grandparents to
by children with autism due to properties that foster tell the cute things their child has said. This study ex-
generalization of skills from teaching conditions to use amined whether incidental teaching could be effective
in children's daily lives (McGee. Krantz, Mason. & in promoting use of social phrases by children with au-
McClannahan, 1983; McGee. Krantz, & McClannahan. tism. Also of interest was whether stimulus control over
1985. 1986). Also relevant to attempts to develop social use of these phrases could be shifted from conditions of
conversations are fmdings that incidental teaching can direct verbal prompting and tangible rewards to situa-
be used to secure verbal initiations and responses that tions that better represent the everyday activities that
enable children with autism to engage in reciprocal play call for social language in young children.
interactions with peers (McGee, Almeida, Feldman.
Sulzer-Azaroff. & Feldman, 1991). However, an essen- Methods
tial feature of incidental teaching is contingent access Participants
to highly preferred rewards, and applications of inci-
Three boys with autism participated. All were en-
dental teaching to children with autism have virtually
rolled in a university-based preschool that was attended
always involved contingent access to highly preferred
by a majority of typical children, and each had been
tangible rewards.
independently diagnosed by experienced professionals
Stimulus fading strategies have shown promise in and met criteria for Autistic Disorder as outlined in the
promoting certain types of social initiations. For exam- Diagnostic and Statistical Manual (American Psychiat-
ple, time delay techniques have been employed as a ric Association, 2001).
means of fading verbal cues for children with autism to When the study began. Child 1 was age 5 years, 2 months
seek information by asking questions (Taylor & Harris, of age, and he had been enrolled in the preschool for
1995). Generalized question-asking and information- 18 months. He had entered the preschool with less than
gathering skills were further developed in young chil- 10 functional words, along with some delayed echolalia,
dren with autism by stimulus fading along dimensions but by the time of the study Child 1 had a large expres-
of reinforcement potency and predictability (Koegel. sive vocabulary and he spoke primarily using phrase
Camarata, Valdez-Menchaca, & Koegel, 1998). speech and brief sentences (mean length of utterance
This present study combined incidental teaching and was 3.52). At the time of the study, Ihe receptive vo-
stimulus fading procedures in an effort to promote use cabulary of Child I was estimated to be at an age equiv-
of age-appropriate social phrases by preschoolers with alent of at least 48 months; he now longer used delayed
autism. Child-culture language includes the colloquial echolalia. but he still had a tendency to repeat phrases
words and phrases associated with current fads (e.g.. several times.
popular phrases from television shows such as Dora Child 2 was 4 years, 11 months when he began par-
or Blue's Clue's), along with more subtle catch phrases ticipation in the study, and he had been enrolled in the
that children often use (e.g.. Wow! or It's mine!). Child- preschool for 10 months. He was mute when first en-
culture language may also be one ofthe criteria by which rolled, and he entered early intervention with a cor-
typical peers choose their playmates, or by which they rected cleft palate and shortened vocal cords. At the
provide social reinforcement to their classmates. Infor- time the study began. Child 2 had been using expressive
mal observations of child-preferred topics implicate age language for only 6 months. Although he had at least
appropriateness as a salient factor in gaining peer ap- 50 meaningful words in his everyday vocabulary, and
proval (e.g., TeleTuhbies are of great interest to young he was capable of combining up to 4 words into com-
toddlers, but paraphernalia such as TeleTiihhie book- plete sentences, he seldom spoke unless directly prompted
bags and lunchboxes are discarded almost immediately to do so. When he did communicate, it was usually with
upon transition into preschool). gestures or one-word utterances (mean length of utter-
In pilot research preceding this investigation, an in- ance was 1.40). Standardized tests suggested that the re-
clusive group of children with autism and typical chil- ceptive vocabulary of Child 2 was at an age equivalent of
dren were taught to use "Knock, knock" and "Why did 31 months; however, formal test results probably under-
the chicken cross the road?" jokes. Teaching conditions estimated his understanding of language because he had
were designed to generate high response rates among learned to read at least some words with comprehension
the typical children to provide abundant models to their even before he was able to speak.
peers with autism. An unanticipated side effect was Child 3 was 4 years, 9 months of age when the study
parental annoyance, which resulted when typical peers began, and he had been enrolled at the preschool in
burdened their family dinner tables with continual rec- which the study was conducted for 15 months. During
114 McGee and Daly

the year preceding the study, the expressive vocabulary room activities while permitting ongoing supervision by
of Child 3 had grown from less than 10 functional words the classroom's lead teacher. The two series of gener-
to more than 75 words; however, much of his language alization probes were conducted in a cordoned-off area
consisted of Immediate or delayed echolalia. occasional (3 X 2 m) of the classroom's firee-play zone. One-way
jargon, and he also engaged in some repetitive speech. observational windows, along with videotapes of all
Child 3 usually spoke in one- or two-word utterances sessions, permitted research supervision from outside
(mean length of utterance was 1.56), and his receptive the classroom.
vocabulary was estimated to be at an age equivalent of
approximately 30 months. Teaehing Materials
Four typical children (three boys and one girl), who A pool of highly preferred toys was selected in the
ranged in ages from 3 years, 5 months to 5 years, 1 month, context of a "school store," which was offered period-
were present throughout all experimental sessions, but ically to provide children with opportunities to sample
ihey received no direct instruction in use of the social and select from an array of manipulative and/or elec-
phrases targeted in this study. The typical peers pre- tronic toys. For purposes of this study, most of the toys
sented with normal intellectual functioning, they had no were newly purchased with the intent of providing choices
evidence of disability, and their language abilities ranged among a wide variety of sensory outputs (e.g.. toys with
from average to above average. Mean length of utterance visual qualities, auditory toys, tactile features). To further
was nearly the same for the three typical boys (3.50,3.72, ensure that the toys would be engaging, after each child
and 3.74 words), and the typical female used slightly selected a toy he or she found to be attractive, the lead
longer verbalizations (mean length of utterance - 4.32). teacher observed and noted whether ihe child played ap-
Verbalizations of the typical children varied from one- propriately with the toy for at least 5 min. Following 5 min
word comments to complex sentences composed of up to of sustained toy engagement, or following no more than
16 words. 2 min of nonengagement, participants were encouraged
The children with autism and typical peers partici- to return to the table to seek a replacement toy. 'ITiis pro-
pated within one of two groups. Child 1 and Child 3 were cess continued until more than a dozen toys had been
assigned to the first group, along with two typical boys. selected that elicited sustained task engagement in at least
Chiid 2 participated in the second group, along with the one of the children, and at least three toys had been
third typical male peer and the typical female peer. identified that elicited sustained task engagement by each
participant. This process was repeated approximately half-
The conversation teacher was a BA-level research as-
way through the study, at which time new toys were added
sistant, who conducted all baseline and incidental teach-
to the pool of previously selected favorites.
ing sessions, as well as retention assessments and setting
generalization probes. Another experienced BA-levei Included among the most frequently selected toys
teacher implemented the classroom's regular free-play used in the study was a talking telephone, a toy movie
manager routine throughout a series of probes designed camera, a wind-up steamboat featuring Disney charac-
to assess generalization across teachers. Both the conver- ters, a Cookie (Monster) Counter computer math game,
sation teacher and the free-play teacher had previously a playhouse with Sesame Street characters, a slinky, a
been provided with systematic training via checklist- kaleidoscope, a stamp and inkpad, a music box, a View
based performance appraisals, and both had mastered Master, an electronic piano, and Hot Wheels cars with
their assigned teaching routines and incidental teaching a toy garage. These same toys were used during all
procedures (cf. McGee & Morrier, 2005). llie conversa- teaching and assessment sessions conducted as part of
tion teacher was also provided with detailed written in- this study, but novelty was preserved by not making the
structions for how to implement experimental protocols, toys available in the classroom at times not associated
and he received daily feedback regarding his adherence with this study (except for during a 10-min period fol-
to the prescribed study methods. lowing free-play generalization probes, during which
classroom children who did not participate in the study
Settings were invited to play with the toys).
Al! experimental sessions were conducted within tra- Prior to inviting the children to a daily tabietop ac-
ditional early childhood activities, which were regularly tivity, a toy that the teacher recognized as one of each
offered in the daily schedule of the preschool in which child's favorites was placed on the table in front of the
the study took place (cf. McGee. Daly. & Jacobs, 1994; child's chair. Extra toys were in a box accessible only to
McGee, Daly, & Morrier, 2001). Baseline and teaching the teacher, but a chiid could exchange his toy for one
sessions, as well as retention assessments, took place at from the teacher's box by simply requesting to do so.
tabletop activities conducted at a kidney-shaped table The conversation teacher aiso volunteered a choice of
in an area (6 x 4 m) that accommodated snack and new toys from the replacement box when a child was
lunch activities at other times of the day. Extraneous observed to be unengaged in either active toy play or
classroom noise was reduced by a glass wall partition social interaction for approximately a minute. When
that physically separated the area from concurrent class- two children simultaneously wanted the same toy, the
 Age-Appropriate Social Phrases 115

classroom's sharing rule governed possession of the toy and language was coded separately for each participant.
(i.e., a disputed toy had to be relinquished after two Verbatim language transcripts were prepared for all
min). During tree-play generalization probes, the toys participating children with autism throughout the study.
were arranged on loy display shelves among other toys In addition, verbatim language transcripts were coded
that were usually available in free-play throughout the for five arbitrarily selected sessions in which each ofthe
school day (e.g.. wooden blocks, puzzles, and regular four typical peers participated.
free-play area manipulative toys, which were routinely The data collection task in coding each transcript was
rotated into the classroom on a weekly basis as older to record every verbalization emitted by a given child
toys were rotated back into storage). throughout the session, exactly as it was said. Although
observation and recording were continuous throughout
Measurement of Dependent Variables
the session, data were entered on data sheets as one
Two social phrases were specifically targeted for teach- utterance per line. An utterance was defined as one or
ing, including AU right and You know what? Minor vari- a group of dietionary-listed words, which usually ex-
ations of each phrase, which were functionally equivalent pressed a complete thought on a given topic. However,
to the original phrase, were also counted as occurrences an arbitrary rule stipulated that pauses of 2 s or longer
of the target phrases (i.e.. Right when used to mean okay ended an utterance, regardless of whether the next ut-
or correct, and Know what'?). Excluded from data analy- terance was on a same or different topic. Specific codes
sis were component words used to convey a different indicated vocal noises (i.e., VN to indicate sounds such
meaning ("right here," "right there." "what is it?," "I know." as animal growls or vehicle noises) and unintelligible
or "you"). vocalizations (UI),
Selection and validation of social phrases Descriptive information on children's language at the
To obtain a pool of age-appropriate social phrases time of entry to the study, along with overall language
that would appeal to both children and adults, we so- progress of the children with autism, was evaluated by
hcited opinions from local area preschool teachers and comparing verbatim language transcripts from five ar-
samples were obtained of everyday language use by bitrarily selected baseline sessions lo five transcripts
typical preschoolers. The teacher questionnaire included from retention assessments conducted during later teach-
multiple choice, "fill-in-the-blank," and open-ended que- ing phases (i.e.. Phase 4 or 5), Differences in the language
ries on "cute" sayings commonly used by typical chil- of participants with and without autism were examined
dren in their classrooms. Fifteen teachers returned by comparing language samples of children with autism
completed questionnaires, and all had selected either to samples obtained for the typical peers. For every block
All right (13 of 15) or You know what? (14 of 15). In- of Eve transcripts that composed each child's language
formal review of typical children's verbatim language sample, data were summarized as total number of words,
transcripts, whieh were being collected as part of a number of different words, mean length of utterances,
concurrent research project in the ciassroom where the number of unintelligible and/or meaningless utter-
study took plaee, verified that most typical children pe- ances, and proportion of overall conversational speech
riodically used one or both of the social phrases selected that was composed of the target phrases.
by local preschool teachers. During preparations for the
study, one child with autism was overheard using the Interobserver agreement
phrase "Put it right here." and "right" was used once Observer agreement was assessed in 25% of each
when a child with autism immediately echoed a peer's child's assessment sessions, and agreement checks were
comment for no apparent reason; otherwise, there were distributed evenly across study sessions. Observers were
neither records nor observations of a child with autism arbitrarily assigned to serve as the primary or reliability
using either the phrase All right or the phrase You know observer for a given session. The independence of pri-
what? mary and reliability observations was ensured by ar-
ranging for two observers to view and score videotaped
segments at different times (e.g., the synchrony of two
Verbatim language transcripts
separate tape viewings was facilitated by audio dubbing
The primary variable of interest was use of either of
a "start" signal onto the beginning of eaeh videotape).
the target social phrases, and the number of phrases
The oeeurrence agreement index, which was computed
used per session was summarized as a rate measure (i.e.,
as the total number of target social phrases scored as
total number of phrases used per session divided by
occurrences by both the primary observer and the sec-
5 min per session). Six research assistants were trained
ondary observer divided by the total number of social
on a verbatim language transcription coding system until
phrases seored by either observer, was .90.
reaching 80% agreement with master training and test
tapes; scoring assignments for each trained observer Experimental Design
were then randomly distributed across every child, con- A multiple baseline aeross participating children with
dition (ie,. baseline or teaching), and type of assessment autism was used to demonstrate the effects of incidental
session. All experimental sessions were videotaped. teaching and stimulus fading on acquisition and use of
 16 McGee and Daly

two social phrases. Baseline sessions were conducted tivity, so that children with autism received twice the
daily in a tabletop activity, during which the target so- level of teacher contacts relative to those provided to
cial phrases were neither prompted nor reinforced. Fol- typical peers. Teacher contacts with typical children
lowing 8 days of baseline assessment, direct instruction were identical to those provided to them during baseline
was introduced for Child 1 in the first group. Child 2 sessions; thus, typical peers were offered general en-
entered teaching in the second group after U days of couragement for verbalizations and task engagement
baseline, and Child 3 joined teaching sessions in group but no specific instruction in use of the target social
one after 17 days of baseline assessment. Teaching ses- phrases. For the children with autism, criteria for moving
sions took place daily during the same tabletop activity from one teaching phase to the next was based on the
in which baseline assessments had been conducted. The child's use of the target phrase at the least intrusive
social phrases were taught across a five-phase teaching prompt available for at least three consecutive teaching
sequence, across which direct prompts to use the phrases opportunities, or throughout at least 75% of opportu-
were gradually faded. Teaching .sessions were preceded nities within a session. Every child at the tabletop activ-
by retention assessments, which were conducted at the ity began the session playing with one of his favorite
same tabletop activity as baseline and teaching sessions, toys.
to measure unprompted use of previously acquired so-
Phase llnitial teaching of "All right"
cial phrases. Generalization across teachers, as well as
To create opportunities in which to teach a child with
generalization aaoss activity/settings (i.e. free-play), was autism to use the phrase All right, the conversation
assessed in biweekly probes throughout baseline and teacher made a brief remark sucb as "Let me have a
teaching. turn" as he gently retrieved the toy. An incidental
teaching episode was initiated when the ehild gestured
Baseline for or requested return of the toy. at which time the
Throughout basehne. children with autism and their teacher looked at him and prompted (first opportunity,
typical peers were invited to participate in one of two with verbal prompt), "You can play with the movie
tabletop activities, according to their assigned group. camera next, all right?" If the child responded by saying
Baseline sessions, which lasted 5 min. began with the All right or Right within 5 s of the teacher's first prompt,
conversation teacher seated across the table from the the teacher praised with confirmation of the response
children as they played with highly preferred toys. (e. g.. "Fantastic! Right!") and provided the child with
The assessment protocol called for the teacher to make access to the desired toy. When the child did not re-
brief contacts with each child, rotating from right to left spond by saying the target social phrase within 5 s of
across the child(ren) with autism and the typical peers. the teacher's model, the teacher more direetly prompt-
Teaching contacts with each child began when the teacher ed (second opportunity, with verbal prompt), "Say, All
"looked and paused" at the child's ongoing play activity, right!." If the child responded by saying the target phrase,
without making a comment. When a child acknowledged the teacher gave the ehild confirming praise and access to
or spoke to the teacher, or following 5 s without an ac- the desired toy. When there was no correct response
knowledgement, the teacher offered a friendly comment within 5 s of the second more direct prompt, the teacher
or "talk-up" about the child's ongoing play acUvity. Typ- made an exit comment ("All right, I'm going to play with
ical children almost always initiated interaction with the Billy now") and then withdrew access to the toy for a
teacher following either a Look/pause or Talk-up. When period of one to two min. The teacher then directed his
a child with autism failed to respond within 5 s of the attention to the next child in the rotation. TTiroughout the
teacher's Talk-up, the teacher attempted to elicit language session, prompts were repeated at each contact with the
by asking a question about the child's ongoing play ac- child with autism as needed to provide as many
tivity (avoiding questions with yes/no answers). Every opportunities as possible for the child to practice saying
child contact ended with the teacher offering general praise the target stxial phrase.
for the child's appropriate engagement and/or verbaliza-
tions (e.g., "You're talking and playing so nicely"), and the Phase 2-Initial teaching of "You know what?"
teacher then moved along to eontact the next child in the As in Phase 1. teaching opportunities were created
rotation. The baseline assessment protocol specifically pre- when the conversation teacher unobtrusively retrieved
cluded the teacher's use of verbal instructions, models, or the child's toy, which the teacher looked at intently as he
praise for use of either target social phrase. played with it. An incidental teaching episode began
when a ehild with autism initiated by gesturing for or
Teaching Sessions requesting the toy. at which moment the teacher in-
Teaching sessions were also eonducted daily for 5 min terrupted the child's request as quickly as possible by
at the same tabletop play activity where basehne as- prompting (first opportunity, with verbal prompt), "Say,
sessments had taken place. The conversation teacher You know what?," The teacher then looked back at
alternated his contacts between a child with autism and and resumed play with the toy. If the child responded
one or the other of two typical peers present at the ac- by saying, "You know what?" within 5 s of the teacher's
 Age-Appropriate Social Phrases 117

first prompt, the teacher praised and confirmed the re- toy without using the social phrase You know what'?,
sponse (e.g.. "Great! What?") and then gave the toy to or when he sat quietly waiting for his turn for 5 s, the
the child. When the child did not respond by saying the teacher prompted, "Say. You know what?." and re-
target social phrase within 5 s of the teacher's model, or sumed playing wilh and looking at the toy (second op-
if the child began to give an echolalic response (e.g., "say portunity, with verbal prompt). If needed to secure the
you know what"), the teacher interrupted the child's child's nonechoed use of the target phrase, the teacher
echoed response as quickly as possible and prompted fell back to the more intrusive Phase 2 prompting. "Say,
(second opportunity, with verbal prompt), "Say. You You know what? I want the toy! Know what?" (third
know what? I want the toy. Know what?." If the child opportunity, with verbal prompt). Children moved into
responded by saying the target social phrase, the teacher Phase 5 after learning to use the target phrase prior to
provided him with confirming praise and access to the
the teacher's model of the phrase.
desired toy. When the child failed to answer the second
prompt with a nonechoic response, the teacher gave an
exit comment ("You know what? I'm going to play with Phase 5 Final fading of cues to use both social phrases
Billy now") and temporarily withdrew the toy and To teach the children to use All right! or You know
teacher attention. The child moved to Phase 3 when he what? in situations in which a child did not have to say
was able to respond at criterion level to the first-level the social phrase to get a toy. Phase 5 incidental teaching
prompt. episodes began while the child had access to his pre-
ferred toy and was playing with it. The teacher's agenda
in Phase 5 was to alternate incidental teaching of the
Phase 3 Prompt fading for "All right"
phrase All right with teaching episodes that targeted the
The conversation teacher again created teaching op-
phrase You know what? However, when a child with
portunities by subtly taking the child's toy and playing
autism used either social phrase at any time during the
with it. An incidental teaching episode began when the
session, the teacher responded wilh an immediate con-
child gestured for or requested return of the toy. at
firmation and permitted the child to control access to the
which time the teacher looked at him and commented
toy for at least 5 s following use of the phrase.
about his own play with the toy (first opportunity, with-
out verbal prompt) (e.g., "i'm making a new design with To set the occasion for teaching increasingly inde-
the stamps."). If the child responded by saying All right pendent use of All right during Phase 5, an incidental
or Right within 5 s of the teacher's comment, the teacher teaching episode began with the teacher's observation
praised with confirmation of the response (e.g.. "Fan- of the child's play activity, along with a comment such as
tastic! Right!") and provided the child with access to "I see that Mickey Mouse enjoys his boat!" (first op-
the desired toy. When the child did not use the target portunity, with no control of toy access and no verbal
social phrase within 5 s of the teacher's first comment. prompt). If the child acknowledged tbe teacher's com-
the teacher fell back into the Phase I prompting se- ment by saying Right or All right, then the teacher
quence by saying (second opportunity, with verbal prompt), praised the child and permitted him to continue playing
"You can play with the toy next, all right?" If the child with the toy. When the child did not acknowledge the
used the target social phrase within 5 s of the teacher's teacher's comment within 5 s, the teacher fell back to
model, he received confirming praise and toy access. the Phase 3 prompting sequence by subtly retrieving
When necessary to ensure that the child used the target the toy and commenting about his own play actions "Let
social phrase, the teacher prompted more directly with me help him drive the boat for a second" (second op-
(third opportunity, with verbal prompt). "Say, All right." portunity, with control of toy access but no verbal
The criterion for mastery was use of the social phrase All prompt). As before, if the child responded within 5 s of
right in response to a teacher's comment that did not the teacher's comment by saying All right or Right, then
include a verbal model, at least 75% of opportunities avail- the response was praised and the toy was quickly re-
able to do so. turned to the child. When the child did not use a social
phrase within 5 s of the teacher's second comment, the
teacher modeled the target social phrase by saying. "You
Phase 4-Prompt fading for "You know what?" can have the boat back in a minute, all right?" (third op-
The conversation teacher quietly retrieved the child's portunity, with control of toy access and verbal prompt).
toy. which he looked at intently as he played with it. By the time any of the participating children reached
When the child gestured for or requested the toy, the Phase 5. they always responded to one of these three
teacher ignored the child's initiation and continued teacher comments by saying the target phrase.
looking at and playing with the toy for a period of up to Prior to teaching episodes aimed at providing in-
5 s (first opportunity, without verbal prompt). If the creasingly independent practice in use of the phrase You
child used the social phrase You know what?, either be- know what?. Xhii child had the toy and was playing with it
fore or within 5 s after initiating for the toy, the teacher when the teacher quietly joined in with tbe child's play
enthusiastically answered {What?) and gave the toy to while looking at the toy (first opportunity, with joint con-
the child, if the child attempted a second time to get the trol of toy access and no verbal prompt). If the child
118 McGee and Daly

used the phrase You know what? within 5 s of when the play action to create an opportunity for teaching (e.g., by
teacher began touching and playing with toy. then the placing a hand on the bridge where the child was about
teacher looked up at the child asking What? and then to roll his toy car). However, teachers did not directly
withdrew his hands from the toy. When the child did not prompt or praise use of the target social phrases during
use the phrase You know what? within 5 s of joint toy generalization probes. The two groups of children par-
play, the teacher quietly took full control of the toy and ticipated separately in designated groups, and each group
watched intently his own play actions as if fascinated participated in two consecutive 5-min generalization
by the toy (second opportunity, with control of toy probes.
access and no verbal prompt). If the child said You
know what? to solicit the teacher's attention and/or to Transfer across teachers
initiate for tbe toy within 5 s of the teacher's toy re- The first half of each generalization probe session was
trieval, then the teacher looked up, answered What?, designed to assess children's transfer of social phrases
and then quickly gave the toy to the child. When the from use with the conversation teacher to use with a
child tried to get the toy without first using the target different teacher who had not been associated with
social phrase, the teacher prompted "Say, You know teaching. Specifically, these probes were conducted by a
what?," and resumed playing with and looking at the toy preschool teacher who was equally experienced as the
(third opportunity, with control of toy access and verbal conversation teacher, but she was not informed about
prompt). By the time a child progressed into Phase 5, the target social phrases.
he nearly always responded before the verbal prompt: Transfer across activities/settings
however, as needed, the verbal model consistently se- The second generalization probe, which followed
cured every child's use of the target social phrase. immediately after the teacher transfer probe, provided
Mastery of all five teaching phases was achieved when for evaluation of a child's ability to transfer use of so-
a child used the target social phrases during joint play cial phrases from the teaching session to use in a less-
with the toy. before verbal models or directions were structured free-play activity held in a different area of
provided by the teacher. After a child met criterion the classroom. The conversation teacher conducted these
on Phase 5. the frequency of instruction was reduced to probes.
one weekly maintenance session (in Phase 5 teaching
conditions). Fidelity of Implementation
The task of tracking the progress of three children
Retention Assessments across a five-phase prompt-fading strategy and tbe vari-
Retention assessments were conducted during a ety of assessment conditions conducted in this study were
5-min tabletop activity on the weekday following each sufficiently complex that it seemed prudent to monitor
teaching session (i.e.. at least 24 hr after most teaching fidelity of implementation. Research assistants reviewed
sessions, and at least 72 hr following Friday teaching teaching session videotapes, which had been previously
sessions); these sessions occurred immediately prior to scored for children's verbatim uses of scKial phrases: com-
the next teaching session. Assessments began with the ponents of incidental teaching episodes were coded as
conversation teacher seated across the table trom the antecedent events (i.e., did the teacher provide a verbal
children, who were provided with highly preferred toys. model, or manipulate the child's toy, or take no action)
The conversation teacher rotated contacts across every and as consequent events (i.e.. teacher praise, response
child participating in the activity. Teacher contacts con- confirmation, provision of toy access, or no action) for
sisted of general praise for talking and task engage- each child's use of a target social phrase. Research assis-
ment, responses to children's queries or comments, and tants recorded narrative notes on data sheets associated
"talk-ups" about a child's ongoing play actions. The re- with retention and generalization assessments; observ-
tention assessment protocol specifically precluded the ers were asked to document any occasion in which the
teacher's provision of verbal prompts, models, or behavior- teacher prompted or rewarded children's use of the target
specific praise for use of either target social phrase. social phrases, as well as to note videotape irregularities.
Free-Play Generalization Probes The most common procedural errors consisted of vid-
Generalization probes, which were scheduled to take eotaped segments that had been recorded for durations
place at least 2 br after teaching sessions, were con- shorter than originally planned; these errors yielded un-
ducted in a free-ptay activity approximately every other derestimates of children's language productions in ap-
week throughout baseline and teaching conditions. The proximately seven sessions. During teaching sessions,
teacher was asked to follow the classroom's regular free- procedural errors occurred primarily when a child met
play teaching routine, which consisted of circulation criterion in a given phase and the phase shift was de-
among all children present while praising toy engage- layed by an extra day. On five occasions during retention
ment and offering incidental teaching of language. Toys assessments, the conversation teacher modeled a target
on the free-play shelves were available to all children, social phrase, and he once praised a child with autism
although a teacher sometimes briefly interrupted a child's for using a target social phrase during a generalization
 Age-Appropriate Social Phrases 119

probe. In addition, although the generalization probe children also transferred use of the social phrases ac-
free-play teacher was uninformed regarding target phrases, quired during teaching sessions to the unprompted con-
she once inadvertently modeled the phrase That's right. ditions of the tabletop retention assessments.
Five of these seven errors were followed by immediate or Figtire 2 illustrates that all children with autism showed
later use of a social phrase, but these prompted responses some transfer in their use of target social phrases from the
were omitted from data summaries (eliminating three so- tabletop teaching sessions to free-play activities. Two of
cial phrases used by Child I during retention assessments, the boys with autism generalized their use of the social
one phrase used by Child 2 on a retention assessment, and phrases from interactions with the original conversation
one phrase tised by Child 3 on a baseline probe). In sum, teacher to interactions with a different teacher.
error corrections ensured that, with the exception of teach- All children achieved criterion for each of the first
ing sessions, all responses represented in Figures 1 and 2 two teaching phases within one or two sessions, but their
were unprompted uses of target social phrases. later acquisition patterns varied from child to child.
Upon introduction into teaching. Child I immediately
began to use the new social phrases during both teach-
Results
ing activities and in the unprompted conditions of the
As shown in Figure 1, there was no use of the target tabletop retention assessments. Child 1 also quickly met
social phrases by any of the three participants during criterion in Phase 3 and in Phase 4 (i.e.. requiring four
baseline, and all demonstrated immediate increases in sessions to master Ail right without a model, and three
use of the target social phrases when teaching was in- sessions to learn how to use You know what? without a
troduced into the tabletop teaching activities. AU three model). Despite rapid progress through the first four

Baseline Incidental Teaching -t- Social Conditioning

2,00+ - / ^ Teaching Sessions ^

1.50-
,:
St ;
1.00-

CD
Acquisition Probes t
01
.50-
Hi
n
.c 0-
a.
75 2.00+ -
c
o
rsat

1.50-
u
o 1.00-
o
teo

.50-
3
C
S 0-

I 2,00+ -
£m
CC 1,50-

1.00-

.50-

0-
i I 1 1] r I ) I I I I I I I I i' I 1 [ I I I I I I I I I I I I I 1 1 M I I I I I I J I 11 I I i' I

5 10 15 20 25 30 35 40 45 50 55

Tabletop Activity Sessions

Figure 1 Rate per minute of target social phrases used at a labletop activity during baseline, during teaching sessions (dashed lines) and
during unprompted reteniion assessments (solid lines).
120 McGee and Daly

Baseline Incidental Teaching + Social Conditioning

2.00+ -

1,50-

1.00-

.50-

0-
a
g 2,00+
o
a 1,50-
i2
0)
o 1,00-
o
"5 .50-
o
"3
c 0-

,? 2.00+-

1.50-

1,00-

.50-

0- X- -

Free-play Generalization Probe Sessions

Figure 2. Rate per minute of target social phrases used in free-play during baseline and teaching with the conversation teacher (sohd
lines) and with a teacher not associated with teaching sessions (dashed lines).

teaching phases. Child 1 needed 19 sessions in Phase 5 by comparing the frequency with which the participants
before he was able to consistently use both target with autism used unprompted social phrases to the fre-
phrases without a verbal model (in conditions in which quency at which the typical peers used these same
he had control of access to his preferred toys). phrases during the same activities. Thus, during un-
Child 2 and Child 3 showed similar patterns of rapid prompted retention assessment conditions, the typical
acquisition of target social phrases during teaching ses- peers used one or the other of the social phrases at an
sions. For example. Child 2 learned to use Alt right with- average rate (responses per minute) of 0.07 (rates for
out a verbal model after two teaching sessions, and he individual typical children were 0. 0.04. 0.12, and 0.12).
learned to use the phrase You know what? without During the same unprompted retention assessments,
a prompt after oniy four teaching sessions. However. Chiid I's average use of the target phrases increased to
Child 2 showed virtually no transfer to unprompted as- the same levels (i.e.. 0.12) at which two of the typical
sessment conditions until nearly 2 weeks into the Phase peers used the phrases. Child 2 and Child 3 actually used
5 teaching condition. Child 3 required only one session the target social phrases at higher rates than the typical
to learn to use the phrase Alt right without a verbal peers, averaging 0.15 and 0.48 responses per minute,
mtxlel. but he struggled across nine sessions before learn- respectively. As prescribed by the teaching protocol, all
ing to use You know what? without a model. He then three children with autism used the social phrases at
moved quickly through Phase 5, which he mastered in even higher levels during teaching sessions than during
only three sessions. the unprompted conditions in retention assessments.
The significance of results indicating generalized use Interestingly, the typical children and boys with au-
of the target social phrases is most easily interpreted tism showed remarkably similar response preferences.
 Age-Appropriate Social Phrases 121

For example, although Chiid 1 and Child 3 learned to uum of most to least intrusiveness (i.e., verbal prompts
use both target phrases without verbai prompting during were graduaiiy removed, and the teacher's controi of
teaching sessions, during unprompted assessment con- access to desired toys was gradually transferred to the
ditions, both boys primariiy used the phrase All right or chiid). However, within a given incidentai teaching ep-
Rif^ht. Two of the three typical boys also used Alt right or isode, stimuius fading occurred across a continuum of
Right but seldom used the phrase You know what?. least to most prompting to provide opportunity for chii-
When Child 2 began to use the social phrases outside of dren to use language as independentiy as possibie. As
teaching sessions, he used both Alt right and You know with many erroriess procedures, procedural complexity
what? at approximately the same rates. The third and may iimit practical appiications.
oidest typicai boy also used You hunv what?. Guess It was clearly easier to promote generalized use of the
what'?. All rightl, and Righil during tabletop retention target social phrases by Child 1, who entered the study
assessment sessions. with the highest level of language functioning, than it
Examination of verbatim language transcripts of the was to expand conversational language in children with
children with autism revealed marked language progress more substantial ianguage delays. Child 1 also devel-
by Child 2 and Child 3 across the course of this study. oped the most varied and flexible use of the target social
During baseline. Child 2 used a total of only 27 words in phrases. For example, during later teaching and assess-
the five-session sample, and he averaged only three dif- ment sessions, he used the phrase Atl right in contexts
feren! words per session. In contrast, throughout five such as "All right everybody, ifs time to go," "That's not
sessions sampled near the end of the sludy. Child 2 used right," '"All right, when the other kids come." SImilariy,
a total of 98 words, and he averaged use of 10 different during teaching sessions, Chiid 1 cleariy demonstrated
words per session. Child 3 showed similar gains in his an ability to use the phrase You know what? for pur-
increase from 41 total words during five baseline ses- poses other than requesting tangible rewards (e.g., "Know
sions to 118 total words in a five-session sample obtained what I have?." "Did you know what?, you can come into
later in the study. At the same time, Chiid 3 improved my house," and "Tlien you know what?, 1 don't have to
from averaging only four different words per baseline play").
session to using an average of \5 different words during Although Child 2 eventually learned to use both
unprompted activities near the end of the study. The social phrases in varied unprompted situations, he con-
language samples of Child 1 actualiy showed a decrease tinued to use them almost exclusively as direct or im-
in the total number of words used, although his mean plied requests for desired objects (e.g., "You know what?
length of utterance was within the ranges of the typical Want Toyiand" to ask for the song piayed by the music
boys during both baseline and teaching. Child 1 in- box; or "Aii right, want Mickey" to request the Disney
creased the average number of different words used per boat). Chiid 3, whose language was at an intermediate
.session from 37 during baseline to 54 near the end of the level relative to other participants, also used the sodal
study. phrases primarily as requests during unprompted situa-
tions (e.g.. "Know what I want?"): however, by the
Discussion study's end. he began to use Atl right as an acknowledg-
Children with autism were able to acquire social phrases ment, as a conversational filler, and even to express lack
during conditions in which prompts and contingent tan- of interest (e.g.. "All right" in response to the teacher
gible reinforcers were available, and they subsequently telling him to sit down, or "All right, all right, all right"
transferred their use of these phrases across extended pe- when the teacher showed him how to operate a non-
riods and to situations in which there were neither prompts preferred toy).
nor rewards for use of the target phrases. All three boys A direct causal relationship cannot be established
with autism also demonstrated some generalized use of between study procedures and the impressive language
the new phrases in a different activity and setting, and growth demonstrated by Child 2 and Child 3 because
two of the three children with autism used the sociai both boys were concurrently participating in other class-
phrases at ieast once in the presence of a teacher who room interventions. Idiosyncratic response characteris-
had not been directiy associated with teaching sessions. tics of these two boys suggest future improvements in
Aneedotai reports from parents of Chiid 1 and Chiid 2. efforts to teach conversational speech to children with
obtained up to 6 months from the end of the study, limited language. For example, it may be most efficient
suggested continued use of the social phrases at home. to begin teaching of conversation skiiis by introducing
The delay shown by Child 2 and Chiid 3 in transfer- a variety of sociai responses, rather than taking this
ring use of social phrases from teaching to unprompted study's approach of targeting a social response and a
assessments was largeiy an artifaet of the prompt-fading social initiation, in addition, as suggested by other re-
procedures, which did not begin to target independent searchers (Wetherby & Prutting, 1984). it would be
responding untii Phase 3. Thus, across the five-phase helpful if stimulus fading steps could be developed to
teaching sequence, prompt dependency was graduaiiy overcome problems associated with echolaiia (i.e.. Child
reduced or eliminated by fading prompts along a contin- 3's echolalia expedited his iearning during phases that
122 McGee and Daly

provided verbai models as prompts, but his repetitive any daily play activity during which toy access can be
speech interfered when prompts were being faded). Fi- briefiy controlled. Experience has shown this format to be
naiiy. Child 2 and Child 3 had a tendency to cycie useful for children at all ianguage leveis (i.e., in at least
through use of both social phrases when they perceived one case, a child's first word was obtained in the course of
that the teacher's nonverbal behavior caiied for a re- the group cheering Yea! during the lunch session). How-
sponse (i.e.. whether the teacher's action was intentional ever, group unison responding seems to have iimited suc-
or directed at securing a specific response): such ob- cess in efforLs to promote soeia! initiations.
servations underiine the need to explore how deficits This study's primary significance is in replication and
in understanding nonverbal cues interact with the de- extension of previous findings that systematic fading
velopment of verbai sociai communication (McGee & procedures can help to transform functional speech into
Morrier. 2003). language that will be useful In a variety of everyday so-
An important advantage of inclusive settings is that cial situations (Koegel et al.. 1998). The current study
the selection of language goais is facilitated by the ease proved incidental teaching to be an effective strategy for
of coliecting information on what typical children ac- teaching children with autism to use social phrases in
tualiy say. Another advantage of inciusion is the po- exchange for tangible rewards. For two of the three boys
tential for observational learning from peers' use of with autism who participated, stimuius fading proce-
conversational phrases (McGee et al.. 1991; Odom & dures were effective in transferring use of the phrase Atl
Strain, 1986: Strain. McGee, & Kohier. 2001). Peer in- right from the controi of direct teacher prompts to the
teractions were not directly targeted in this study, and control of setting events that call for acknowledgment
it could be argued that the presence of typical peers or confirmation of an agreement (All right). Conclu-
actuaily diminished instructional potentiai by reducing sions are more tentative as to whether a strategy of shift-
the time available for teacher interactions with children ing teacher control of consequent events (i.e., tangibie
with autism. However, anecdotal review of the video- rewards) to the control of children themselves helps to
tapes reveaied that all three children with autism ben- convert a child's need to solicit another person's attention
efited from participating aiong with their typicai peers. (You know what?) to a desire to soiicit another person's
Ironically, the typical peers in Group 1 mistakenly
attention. Overail, this study adds evidence that stimulus
thought lhat the social phrase was "okay," and they peri-
fading techniques may be used to push the boundaries
odically initiated to the teacher and to other children in
of the kinds of communication that children with autism
their group by playfully suggesting it was time to play the
can be taught to use in socially meaningful ways.
"okay game": for whatever reason, both Child I and
Child 3 learned to use /^//r/g/trasacomment meaning the
same thing as "okay." Near the end of the study, without References
any direct instruction in social conversation with peers.
Child I began to use his new sociai phrases to respond to American Psychiatric Association. (2001). Diagnostic and Sta-
MKial bids from peers (e.g.. a peer holding a toy tele- tistical Manual (4th ed.. rev.), Washington. DC: Author.
phone told Child I to "Call me," and Child 1 responded Charlop, M, H., & Walsh, M. E. (1986), Increasing autistic chil-
dren's spontaneous verbalizations of affection: An assess-
"All right, ri! call you"). He also learned to use the new ment of time delay and peer modeling procedures. Journal of
phrases to initiate interactions with peers ("Billy, we can Applied Behavior Analysis. 19. 307-314.
both share, right?"). During one free-play session super- Guess, D.. Sailor, W., & Baer. D. (1976). Functional speech and
vised by the teacher who had not been associated with language training for the severely handicapped. Lawrence,
KS: H & H Enterprises.
teaching sessions, the typical femaie peer from Group 2
Haring, T. G., & Breen. K. (1992). A peer mediated social net-
offered Chiid 2 a coupie of incidentai teaching episodes work intervention to enhance the sociai integration of per-
that were virtually identical to those provided by the con- sons with moderate to severe disabilities. Journal of Applied
versation teacher during daily teaching sessions. Future Behavior Analysis, 25, 265-280.
research is caiied for to determine whether use of age- Hart, B. M,, & Rislcy. T. R. (1968), Establishing the use of
appropriate, child-cuiture language does in fact increase a descriptive adjectives in the spontaneous speech of disad-
vantaged preschool children. Journat of Applied Behavior
child's attractiveness to peers. Analysis, 1, 109-120.
Koegel. L, K., Camarata, S. M,. Valdez-Menchaca, M., &
A practical manner of implementing the reported inter- Koegel, R. L, (1998), Setting generalization of question-
vention in a preschool classroom is to target a "lunch asking by children with autism, American Joumal on Menial
Retardation, 102. 346-357.
word" or phrase per week (e.g., saying Coot or Yes.''whi\e
Krantz, P. J,. Zalenski, S.. Hall, L, J., Fenske, E. C , &
puUing fist downward). The momentum afforded by get- McClannahan, L. E, (1981). Teaching complex language to
ting an inclusive group of children to say the word or autislic children. Analysis and Intervention in Developmental
phrase frequently is often sufficient to yield acquisition Dis(ihilitiex,L 259-291.
by chiidren who are socialiy responsive to aduits, or by McEvoy, M. A,.Nordquist, V. M.,Twardosz,S,, Hcckaman. K. A,,
Wehby, J. H., & Denny, R. K, (1988). Promoting autistic
those who have learned to imitate their peers (McGee children's peer interaction in an integrated early childhood
et ai.. 2001). As needed, the compiex teaehing sequence selling using affection activities. Journal of Applied Behavior
used in this study may be impiemented in the context of Analysis, 21, l'i^-200.
 Age-Appropriate Social Phrases 123

McGee, G. G., Almeida. M. C , Sulzer-Azaroff. B., & Feldman, McGee. G. G.. & Morrier, M. J. (2005), Preparation of autism
R, S. (1991). Promoting reciprocal interactions via peer inci- specialists. In E R. Volkmar. R. Paul. A. Klin. & D. J. Cohen
dental teaching. Journal of Applied Behavior Analysis, 25, (Ed^). Handbook of autism and pervasive developmental dis-
117-126. orders: Volume 2. Asses.sment. interventions, & policy (3rded..
McGee, G. Q . Daly. T . & Jacobs, H. A. (1994), The Walden pre- pp. 1123-1160). Hoboken, NJ: Wiley.
school. In S. L, Harris & J. S. Handleman (Eds.). Preschool Neef, N. A., Walters, J.. & Egel. A, L. (1984). Establishing
education prof-ranis for children wilh autism (pp, 127-162), generative yes/no responses in developmentally delayed chil-
Austin. TX: Pro-Ed. dren. Journal of Applied Behavior Analysi.s. 17. 453-460.
McGee. G. G, Daly, T . & Morrier. M. J. (2001). Walden early Odom, S. L.. & Strain, P. S. (1986). A comparison of peer ini-
childhood program. In J. S, Handleman & S. L. Harris (Eds.). tiation and teacher antecedent interventions for promoting
Preschool education programs for children with autism reciprocal social interaction of autistic preschoolers. Journal
(2nd ed., pp, I."57-190). Austin. TX: Pro-Ed. of Applied Behavior Analysis. 19, 59-71,
McGee, G, G . Krantz. P J.. Mason. D.. & McClannahan. L. E. Schreibman. L. (1988), Autism. Newbury Park, CA: Sage Pub-
(1983). A modified incidental-teaching procedure for autistic lications, Inc,
youth: Acquisition and generalization of receptive object Skinner. B. F. (1957). Verbal behavior. Englewood Cliffs, NJ:
labels. Journal of Applied Behavior Analysis, 16, 329-338, Prentice-Hall. Inc.
McGee. G. G., Krantz, P, J.. & McClannahan. L. E, (1984), Strain. P S,. McGee. G. G.. & Kohler, F W. (2001). Inclusion
Conversational skills for autistic adolescents: Teaching as- of children with autism in early intervention settings: An
sertiveness in naturalistic game settings. Journal of Autism examination uf rationale, myths, and procedures. In M. J.
and Developmental Disorders. 14. 319-.'i30, Guralnick (Ed,). Early childhood inclusion: Focus on change
McGee. G. G.. Krantz. P J.. & McClannahan. L. E. (1985). The (pp. 337-363), NY: Paul Brookes.
facilitative effects of incidental teaching on preposition use Taylor. B., & Harris. S. (1995). Teaching children with autism to
by autistic children. Journal of Applied Behavior Amdysis. 18. seek information: Acquisition of novel information and gen-
17-31. eralization of responding. Journal of Applied Behavior Anal-
McGee, G. G , Krantz. P J., & McClannahan. L. E. (1986). An ex- ysis, 28.3-U.
tension of incidental teaching to reading instruction for autistic Wethcrby, A. M.. & Prutting. C. A, (1984). Profiles of commu-
children. Journal of Applied Behavior Analysis. 19. 147-157, nicative and cognitive-social abilities in autistic children.
McGee. G. G,. & Morrier, M. J. (2003). Clinical implications of American Speech-Language-Hearing Association. 27,364-377,
research in nonverbal behavior of children with autism. In
P Philipot, E. J, Coats. & R. S. Feldman (Eds.). Nonverbal Received: February 27. 2007
behavior in clinical settings (pp. 287-317). NY: Oxford Final Acceptance: May 17. 2007
University Press. Editor in Charge: Lynn Koegei
View publication stats