Behaviour Change

http://journals.cambridge.org/BEC

Additional services for Behaviour Change:

Email alerts: Click here

Subscriptions: Click here
Commercial reprints: Click here
Terms of use : Click here

Functional Communication and Other Concomitant

Behavior Change Following PECS Training: A Case
Study

Angelika Anderson, Dennis W. Moore and Therese Bourne

Behaviour Change / Volume 24 / Issue 03 / August 2007, pp 173 - 181

DOI: 10.1375/bech.24.3.173, Published online: 22 February 2012

Link to this article: http://journals.cambridge.org/abstract_S0813483900002060

How to cite this article:

Angelika Anderson, Dennis W. Moore and Therese Bourne (2007). Functional
Communication and Other Concomitant Behavior Change Following PECS Training: A Case
Study. Behaviour Change, 24, pp 173-181 doi:10.1375/bech.24.3.173

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/BEC, IP address: 128.122.253.228 on 24 Apr 2015

 Case Reports and
Shorter Communications
Functional Communication and
Other Concomitant Behavior Change
Following PECS Training: A Case Study
Angelika Anderson and Dennis W. Moore
Monash University, Australia

Therese Bourne
IDEA Specialist Services, Auckland, and The University of Auckland, New Zealand

The Picture Exchange Communication System (PECS) is widely used to teach chil-
dren with language delays, including those with autism, functional language. A fea-
ture of PECS is that it incorporates principles deemed by some to be pivotal, leading
to broader behaviour change. In this study, a 6-year-old child with autism was taught
functional language using PECS. Along with measures of language gains, concomitant
changes in nontargeted behaviours (play and TV viewing) following PECS training
were observed. Results show increases in manding, initiations and cumulative word
counts, as well as positive changes in the nontargeted behaviours.

T he Picture Exchange Communication System (PECS) is widely used to teach

functional language to children with language delays, including those with
autism, yet there are relatively few studies documenting its efficacy (Charlop-
Christy, Carpenter, Le, LeBlanc, & Kellet, 2002; Kravits, Kamps, Kemmerer, &
Potucek, 2002; Liddle, 2001; Schwartz, Garfinkle, & Bauer, 1998).
A feature of PECS is that it teaches the child to initiate interactions, a
behaviour class deemed by some to be pivotal (Koegel, Carter, & Koegel, 2003),
leading to broader behaviour change and generalisation. Reported collateral
behaviour changes include the acquisition of verbal language. Schwarts, Garfinkel
and Bauer (1998) demonstrated that PECS was acquired successfully by a number of
children with various developmental disabilities, and reported generalisation of
PECS use to untrained settings and concomitant effects on untrained language
functions. Some of the children in that study, including children with autism, also
acquired verbal language. Charlop-Christy et al. (2002) in a study in which three 173
children with autism were taught functional communication using PECS, have also
demonstrated collateral behaviour change, including the acquisition of verbal lan-
guage, but they did not report requests and other initiations separately. Finally, Frea,

Address for correspondence: Angelika Anderson, Faculty of Education, Monash University, Clayton VIC
3800, Australia. E-mail: [email protected]

Behaviour Change | Volume 24 | Number 3 | 2007 | pp. 173–181

 Angelika Anderson, Dennis W. Moore and Therese Bourne

Arnold and Vittimberga (2001) report concomitant behaviour change (reduction in

aggression and increase in choice-making behaviour) following PECS training.
While all these studies report some generalisation results, all data was collected
either in early intervention centres, clinics, or integrated preschools. None of the
studies were carried out in the home, nor did they report generalisation data collected
in home settings. The current investigation provides additional empirical data on the
effectiveness of PECS training with children with autism in a child’s home. The explicit
focus of this study was on concomitant change in behaviours not directly targeted by
the intervention. Our hypothesis was that following PECS training there would be ben-
eficial changes in other behaviours, including increased initiations other than mands,
increased play, and decreased levels of stereotypic behaviour (specifically TV viewing).

Method
The aim of the study was to use PECS to teach functional communication to a
6-year-old boy with autism and to monitor concomitant changes in nontargeted
behaviours (play, television viewing and verbal initiations other than mands). The
study was conducted in the child’s home and all reported language related data was
collected in a nontraining generalisation setting (free-play following instruction).

Participant
The participant was a 6-year-old Samoan boy (pseudonym Todd), diagnosed as autistic
at age 3. Todd scored 42 on the Childhood Autism Rating Scale (CARS) (Schopler,
Reichler, & Rochen Renner, 1988), rating him severely autistic. At the time of this
study Todd’s receptive language was equivalent to that of a 1-year, 9-month-old as
assessed with the Peabody Picture Vocabulary Test — 3rd Edition (Dunn & Dunn,
1997). He lived at home with his family and attended the local school. In a previous,
unrelated study Todd was reported to have no functional language, though he displayed
some echolalic speech. This consisted mainly of jingles and lines from TV commercials
and Todd’s favourite DVD shows. Todd spent much of his time engaged in repetitive
tapping/drumming of a pair of identical objects and/or watching television, to the
extent that he was inaccessible for instruction or other social interactions (Godfrey,
Moore, Fletcher-Flinn, Anderson, & Birkin, 2002).

Settings
All training and observations were conducted in Todd’s home. The living areas in
this home were open plan, with only partial separation of the kitchen and living
room spaces. Training sessions were conducted at the kitchen table, and the free-
play (generalisation) sessions occurred in the living room. Sessions were conducted
between the hours of 4 pm and 6 pm for no longer than 1 hour each.

Intervention
174
The materials used were the PECS communications folder containing from 25 to 30
symbols, a video camera, table, chairs for Todd, the trainer and, in step 1, for the
helper, and an array of toys (details from authors on request).
The communication folder was a green A5 folder containing three pages that
had three stripes of Velcro per page, placed horizontally. This was the same for the
inside front and back covers. On the front of the folder were four stripes of Velcro,
three were placed horizontally and one was placed vertically. One horizontal stripe

Behaviour Change
 PECS Training: A Case Study

had a removable Velcro (catch) strip/card on it that ran the length of the stripe,
and initially included a square with the words ‘I want’ on it. This was the sentence
strip where Todd would make his sentence of ‘I want’ and ‘chips’, for example.
Inside the folder, were 25 to 30 2.5 cm-square colour and black-and-white picture
cards. Each of these represented a preferred activity or food, for example, chips,
block (for lego), cup (for drink).

Research Design
An A, B, C, D design was used with a focus on generalisation effects. All outcome
data was obtained in a no-treatment generalisation condition immediately following
the training.

Procedure
There were four phases in all: Phase 1 was a baseline consisting of observations only
(B 1), Phase 2 was an assessment of readiness for the intervention (B 2), Phase 3
involved training skills necessary for intervention (compliance training); and Phase
4 was the PECS training phase. During all phases other than Phase 1 (which was
entirely free-play) the sessions consisted of 30 minutes of instructional time fol-
lowed by 30 minutes of free-play. During the free-play time the researcher was pre-
sent and available, but did not provide any instruction. The child was free to engage
in any activity he chose, including watching television.

Phase 1 — Baseline 1
This phase was an unobtrusive observation period with no instructional demands or
other communicative initiations from the researcher.

Phase 2 — Baseline 2: Readiness Assessment

This phase was designed to assess Todd’s ability to comply with requests. Working in
the child’s home, an open-plan living space with siblings and parents present at
times, necessitated a degree of compliance by the child to allow any interactions
with the researcher to occur. The requests (turn off the television, come to the
table, sit down, and play) were presented at least 5 consecutive times without pro-
viding any consequences for compliance or noncompliance. Criterion (compliance
within 5 seconds of request on five consecutive occasions) was not met.

Phase 3 — Compliance Training

Assessments from Phases 1 and 2 indicated that compliance training was needed
before PECS training could occur. During this phase Todd was taught to comply with
the following 3-step request sequence using a forward chaining procedure: (1) TV off
now (2) Come, sit at the table (3) Let’s play (details available from the authors).
175
Phase 4 — Training (PECS)
PECS was implemented as described previously (Bondy & Frost, 1994; Charlop-
Christy et al., 2002), up to Step 4 — sentence strip. This included a reinforcement
assessment, initial physical exchange of pictures of preferred items, generalising the
use of PECS by working at and away from the workstation, increasing the array of
choices offered at a time, and encouraging independent use of PECS to mand.

Behaviour Change
 Angelika Anderson, Dennis W. Moore and Therese Bourne

Step 1. The physical exchange included assessment of reinforcers (finding out what
the child preferred in food, activity and objects/toys), fully assisted picture exchange
for pictured item (hand over hand exchange, assisted by a ‘helper’, not the trainer)
of a picture card to the communicative partner, fading of assistance by the helper for
the exchange, and fading of open handed cue (used to prompt a response).
Step 2. Expanding spontaneity included attaching a ‘preferred item’ picture to a
communication board (using Velcro) and then increasing the distance between
trainer and student and the distance between student and board, so that the learner
had to come and ‘nag’ the adult with his mands.
Step 3. Picture discrimination included (1) discrimination training, (2) correspon-
dence checks, (3) reduction of picture-card size.
Step 4. Sentence structure: (1) stationary ‘I want’ picture — child put together a sen-
tence of ‘I want’ plus desired item on the sentence strip, then gave the sentence strip
to the trainer. The trainer was seated at the workstation at the time; (2) moving ‘I
want’ picture — child made sentence strip with ‘I want’ plus desired item. The trainer
was no longer seated at the workstation but away from the child, so that the child was
required to get out of his seat and go to the trainer, and get her attention to place the
sentence strip in the trainer’s hand; (3) Referents not in sight — the learner was able
to construct the sentence strip ‘I want’ plus desired item manding for things not
offered as a choice on their communication board, for example, ‘I want the park’.
During initial stages of the intervention Todd was given access to three symbols:
swing, drink and burger rings. Access to other symbols was gradually increased. By
the end of the study he had acquired the following PECS vocabulary: I want, swing,
monsters, chocolate, chips, lego, twisties, warehouse, mini-M&M’s, playstation,
peaches, bubbles, slinky, play-doh, tickle, car, peaches, koosh ball, apple, banana,
chicken, KFC, toilet, television, drive.
As the sessions progressed all choices were arranged throughout the book ran-
domly. Todd learned to initiate communication in the first three phases of PECS by
choosing his reinforcer and matching picture card, and giving it to his communica-
tive partner. In the fourth phase of PECS Todd was able to initiate communication
by putting his sentence strip together (‘I want’ plus whatever his choice was) and
then take the strip with his constructed sentence and give this to his communica-
tive partner (trainer). Since Todd exhibited echolalic speech at the start of the
study he was encouraged to voice all mands during the exchange from the outset.
During the free-play sessions the PECS materials continued to be available but only
verbal language with or without the use of PECS was recorded.

System of Observation
All sessions were videotaped. For each session a 10-minute segment of free-play was
176 analysed, beginning five minutes after termination of training.

Interobserver Agreement
Interobserver agreement was obtained for all dependent variables in at least 20% of
the sessions in each phase and calculated using the following formula:
Agreements × 100
Agreements + Disagreements

Behaviour Change
 PECS Training: A Case Study

The mean agreement for these sessions was 96%.

Dependent variables
Three behaviour classes were monitored throughout the study: language, play, and
television watching.

Operational Definitions
Language. Language was defined as verbalisations by Todd that could be under-
stood by most people, including those who did not know him well. Subcategories of
language that were specifically recorded were mands, initiations, and new words
used. All language measures were monitored using event recording.
Manding. Any occurrence of Todd asking for an activity (e.g., swing), object (e.g.,
lego) or place (e.g., garden), using a prefix like ‘I want’, ‘Get me’, ‘Give me’, ‘Get’,
‘Give’.
Initiations (other than mands). Any verbalisation Todd directed at another person
present, with or without the use of PECS, that was not a mand and not a response
to another person’s verbal or nonverbal initiation.
New words. These were recognisable words spoken by Todd that had not previously
been recorded. These were recorded and added to a cumulative word list.

Play
Interactive manipulation of an instrument (toys, game, equipment) in the way it
was designed to be used (i.e., building with blocks, pushing a car around, playing
playstation), or interactive contact with another person (specifically tickling). Play
was scored from the moment Todd made appropriate contact with a toy until disen-
gaged, defined as noncontact with all toys for 5 seconds or more. Cumulative dura-
tion was recorded, in minutes, using the counter on the video.

Television Watching
This behaviour was defined as looking at the television for longer than 30 seconds
continuously without interruptions exceeding 30 seconds. A cumulative duration
was recorded, in minutes, using the counter on the video.

Results
Figure 1 displays the number of mands, verbal initiations and a cumulative count of
new words spoken by Todd across all phases in the free-play (generalisation) setting.
No mands were observed in the phases preceding PECS training, whereupon mand-
ing occurred on 12 of 17 days (range 0–22, mean = 4.2). Similarly, no instances of
verbal initiation other than mands were observed during either baseline 1 or 2.
Todd was observed to initiate 7 times on the first day of compliance training and
177
again on 1 occasion on each of the final 2 days of this phase. These were sponta-
neous verbalisations and may have been words from his echolalic repertoire.
Associated with PECS training was an increase in the number of days on which
verbal initiations occurred and in the rate of initiations each day (range 0–24, mean
level = 6.1). The cumulative record of new words spoken shows that Todd was
observed to say a single new word in baseline 1 and 16 words on the first day of
compliance training. These 16 words were a single phrase of echolalic speech of a

Behaviour Change
 Angelika Anderson, Dennis W. Moore and Therese Bourne

FIGURE 1
Number of mands, initiations and the cumulative record of new words for a 6-year-old boy with
178 autism during baseline, compliance training and PECS training in a generalisation setting.

jingle from a TV show. No other new words were recorded until the first day of
PECS training. However following the start of PECS training new words were
observed on 15 of the 17 days, and Todd’s observed spoken vocabulary expanded to
89 words.

Behaviour Change
 PECS Training: A Case Study

FIGURE 2
Time spent watching TV and playing for a 6 year old boy with autism during baseline,
compliance training, and PECS training in a generalisation setting.

Figure 2 displays concomitant changes in the proportion of time Todd engaged in

TV watching and play in the free-play (generalisation) setting across all phases. TV
watc hing was a high probability activity in both baseline 1 and 2 (mean levels 75%
and 91% respectively). The introduction of compliance training was associated with a
reduction in the time Todd spent watching TV (range 0–65%, mean = 28%), and
during PECS training TV watching occurred on only two of the 17 days (range 0–42%,
mean = 3.8%). An almost reciprocal change was observed in the proportion of time
Todd engaged in play behaviour. Zero levels of play were observed in both baseline
phases, with some play being recorded on two of the four observed sessions during com- 179
pliance training (mean = 26%). A further change was observed during PECS training
with Todd engaging in play on 14 of the 17 days (range 0–100%, mean = 45%).

Discussion
These results confirm previous findings that PECS is acquired easily by a child
with developmental disabilities (Charlop-Christy et al., 2002; Frea et al., 2001;

Behaviour Change
 Angelika Anderson, Dennis W. Moore and Therese Bourne

Schwartz et al., 1998), and support previous findings that PECS acquisition is
associated with concomitant behaviour change, including the acquisition of
verbal language (Charlop-Christy et al., 2002; Frea et al., 2001; Schwartz et al.,
1998). Teaching the child to mand was associated with an increase in other, func-
tionally different verbal initiations, as well as a reduction in television watching
and an increase in play. The observed collateral behaviour change or generalisa-
tion adds strength to the argument that PECS targets pivotal behaviours.
This study also demonstrated that PECS training can be successfully imple-
mented entirely in a natural environment (here, the child’s home), and that the
positive changes generalise to a nontreatment setting.
Some limitations of the current study are noted. The focus was entirely on
the effects of PECS training on concomitant behaviour change. Consequently,
data of the actual effectiveness of the PECs training in the training setting were
not collected. Furthermore, carrying out this study entirely in the natural envi-
ronment, with no attempts to control this environment (e.g., other family mem-
bers were present at times, the TV set was not removed or locked) meant that
some initial compliance training was necessary to increase Todd’s accessibility to
instruction/interaction. The compliance training is a possible confound, and it
is possible that at least some observed concomitant behaviour change was the
result of the compliance training alone. However, the television set was always
available in the free-play time period, and the child could have switched it on at
any time in this condition. This illustrates part of the challenge of working
entirely in natural settings. Further exploration of behaviour changes as a result
of targeting pivotal behaviours, particularly with children with autism, in natu-
ral settings is warranted.

References
Bondy, A.S., & Frost, L.A. (1994). The Picture Exchange Communication System. Focus on
Autistic Behavior, 9(3), 1–19.
Charlop-Christy, M.H., Carpenter, M., Le, L., LeBlanc, L.A., & Kellet, K. (2002). Using the
picture exchange communication system (PECS) with children with autism: Assessment
of PECS acquisition, speech, social–communicative behavior, and problem behavior.
Journal of Applied Behavior Analysis, 35(3), 213–231.
Dunn, L.M., & Dunn, L.M. (1997). Peabody Picture Vocabulary Test (3rd ed.). Circle Pines,
MN: American Guidance Service.
Frea, W.D., Arnold, C.L., & Vittimberga, G.L. (2001). A demonstration of the effects of aug-
mentative communication on the extreme aggressive behavior of a child with autism
within an integrated preschool setting. Journal of Positive Behavior Interventions, 3(4),
194–198.
Godfrey, R., Moore, D.W., Fletcher-Flinn, C., Anderson, A., & Birkin, C. (2002). An evalua-
tion of some programmes for children with autistic spectrum disorder in Auckland: Opportunities,
contingencies, and illusions (Report prepared for The Ministry of Education. Auckland
180 UniServices Limited). Auckland, New Zealand: The University of Auckland.
Koegel, L.K., Carter, C.M., & Koegel, R.L. (2003). Teaching children with autism self-initia-
tions as a pivotal response. Topics in Language Disorders, 23(2), 134–145.
Kravits, T.R., Kamps, D.M., Kemmerer, K., & Potucek, J. (2002). Brief report: Increasing com-
munication skills for an elementary-aged student with autism using the picture exchange
communication system. Journal of Autism & Developmental Disorders, 32(3), 225–230.
Liddle, K. (2001). Implementing the Picture Exchange Communication System (PECS).
International Journal of Language & Communication Disorders, 36(Suppl), 391–395.

Behaviour Change
 PECS Training: A Case Study

Schopler, E., Reichler, R.J., & Rochen Renner, B. (1988). The childhood autism rating scale
(CARS) for diagnostic screening and classification of autism. Los Angelea, CA: Western
Psychological Services.
Schwartz, I.S., Garfinkle, A.N., & Bauer, J. (1998). The picture exchange communication
system — Communicative outcomes for young children with disabilities. Topics in Early
Childhood Special Education, 18(3), 144–159.

181

Behaviour Change