LSHSS
Facilitating Peer-Group Entry in
Kindergartners With Impairments in
Social Communication
Jill Selber Beilinson
Lesley B. Olswang
University of Washington, Seattle
V
arious studies have shown that children with
language impairments exhibit deficits in the
area of social communication (Brinton & Fujiki,
1997; Craig & Washington, 1993; Guralnick, Connor,
Hammond, Gottman, & Kinnish, 1996; Redmond & Rice,
1998, 2002; Rice, Sell, & Hadley, 1991). Studies documenting the social difficulties of children with communication
deficits, including children with autism and specific language
impairment (SLI), point to specific difficulties initiating play
and entering peer groups (Craig, 1993; Craig & Washington,
1993; Koegel, 2000; Odom & Strain, 1986; Rice et al.,
1991; Roth & Clark, 1987), which make it difficult for
further opportunities for social interaction.
ENTERING THE PEER GROUP
Before children can engage in a peer interaction, they
first have to find a way to initiate and be accepted by their
ABSTRACT: Purpose: This series of case studies examined
the efficacy of intervention designed to teach peer-group
entry skills to kindergartners with social interaction and
communication deficits.
Method: The participants were 3 kindergartners at the
University of Washington Experimental Educational Unit
(EEU) who were selected because of difficulty with peergroup entry and cooperative play as compared to other
children in the classroom. The intervention program
included direct treatment of the children by the primary
researcher and teachers in the classroom. The intervention
was modeled on research describing a sequential peer-entry
hierarchy that incorporated the children moving from lowrisk strategies to high-risk strategies. Specifically, the
treatment focused on teaching the children to use props to
facilitate the production of high-risk verbal statements.
Results: Results demonstrated increases in (a) children’s use
154
LANGUAGE, SPEECH,
AND
HEARING SERVICES
IN
peers. Several researchers have described successful peergroup entry as consisting of three factors (Dodge, Schlundt,
Schocken, & Delugach, 1983; Putallaz & Gottman, 1981).
First, children use a string of combined tactics to gain
entry into a peer group. Second, successful tactics involve
some amount of risk in being rejected by the peer group.
For example, children who are successful at entering a peer
group indicate their attempts by using low-risk tactics, such
as waiting and hovering or engaging in parallel play, then
moving to using higher risk tactics, such as making a
verbal statement or asking a question to the peer group.
The third peer-group entry factor is entering the group
without drawing too much attention to one’s self by
causing minimal disruption to the peer group.
Dodge et al. (1983) hypothesized that some children
with social communication deficits may not naturally learn
the necessary progression from using low-risk tactics to
using high-risk tactics. Descriptions of children’s entry
behaviors in previous studies (Brinton & Fujiki, 1997;
of props and verbal statements to enter peer groups, (b)
cooperative play, and (c) time spent interacting with peers.
Results also indicated that following treatment, the
children’s behaviors more closely resembled those of their
comparison peers.
Clinical Implications: Results are encouraging for suggesting
strategies for working with kindergartners who exhibit
social communication interaction problems. The data
indicate that a combined speech-language pathologist/
teacher intervention using modeling and prompting with
visual stimuli may be successful in teaching children to use
props and specific verbal statements as a means of entering
peer groups and engaging in cooperative play.
KEY WORDS: peer-group entry, prop use, social communication, classroom intervention
SCHOOLS • Vol. 34 • 154–166 • April 2003 © American Speech-Language-Hearing Association
0161–1461/03/3402–0154
Craig & Washington, 1993; Rice et al., 1991) indicate that
the children with SLI may plateau at a stage where they
use only low-risk strategies. Data suggest that children with
language impairments lack peer-directed behaviors during
the peer-group entry process (Rice et al., 1991; Roth &
Clark, 1987). The children often walk around the room,
watch the partner, attend to nonplay objects, or bang and
throw objects (Roth & Clark, 1987).
Not only do children with language impairments seem to
lack the social skills of peer-group entry, but they also
appear to lack linguistic skills important for entering
groups. Results from a study performed by Craig and
Washington (1993) showed that children with SLI had
difficulty verbally initiating interactions with their peers.
Brinton and Fujiki (1997) found similar findings indicating
that children with language impairments were not able, or
required more time, to enter peer interactions.
Redmond and Rice (1998) described several behaviors of
children with SLI that appear to compensate for their
linguistic deficits in social situations. These behaviors,
which appear in the extant literature, include low levels of
peer-directed behaviors, high levels of adult-directed
behaviors, lower rates of talking, and lower rates of being
talked to by peers (Brinton & Fujiki, 1997; Craig &
Washington, 1993; Rice et al., 1991; Roth & Clark, 1987).
The clinical implications of this research indicate the need
for treatment to focus on the language deficit as a means
of improving social interactions (Redmond & Rice, 1998).
Hadley and Schuele (1998) noted that because the unique
demands of peer-group entry interactions can burden a
child’s verbal abilities, peer-group entry interventions must
also provide linguistic support. From the perspective of
providing an intervention that emphasizes language, Hadley
and Schuele suggested several specific components that
should be included in such a language-based approach.
INTERVENTION
Hadley and Schuele (1998) suggested an intervention
plan based on work with preschoolers with SLI who had
difficulty entering peer groups and sustaining play. This
intervention focused on three essential components: (a) the
child is assigned a role to establish him or her as part of
the group, (b) the child is prompted by an adult to initiate
interactions, and (c) the child is given a highly valued prop
(toy) for the purpose of integrating him or her into the peer
group. Hadley and Schuele argued that these components
could provide linguistic support that might help make the
peer-group entry process easier. Finally, the authors
emphasized the importance of the natural setting, like the
classroom, for treatments that focus on social communication. This component serves to increase generalization and
ecological validity.
The purpose of this project was to examine the efficacy
of a peer-group entry intervention designed to address the
components specified by Hadley and Schuele (1998) with
children with language impairments. The treatment package
in this study included instructing teachers to prompt
children to use props (toys) to facilitate the production of
high-risk peer-group entry behaviors. To begin to address
the efficacy of the intervention, the following questions
were asked:
• Will the intervention increase the children’s use of
high-risk behaviors and decrease their use of low-risk
behaviors?
• Will the intervention increase the children’s use of
props during peer-group entry?
• Will the intervention increase the children’s amount of
time in cooperative play and decrease the amount of
solitary play?
METHOD
Participants
This project included 3 kindergarten children (ages 5:6
[years:months]–6:3) as case studies. The children were all
enrolled in a full-day integrated kindergarten classroom at
the University of Washington Experimental Education Unit
(EEU) and were selected by their teachers for participation.
All of the children were described as exhibiting difficulties
in the area of peer interaction as compared to their
classmates, specifically in peer-group entry and group play.
Teachers noted that these target children spent the majority
of time playing alone or had difficulty initiating interactions with classroom peers without causing a disruption. In
addition, these children had individualized education plan
(IEP) goals targeting peer interaction and cooperative play.
The target children also met the following inclusion
criteria: (a) native English speakers, (b) no diagnosis of a
behavioral disorder as reported by parents and teachers, and
(c) no report of a hearing or vision problem as reported by
parents or teachers. Although test results were not used to
qualify the children for this study, information from
standardized measures helped to determine the appropriate
language level to use with each child during treatment. The
Test of Language Development Primary–3 (TOLD:P–3;
Newcomer & Hammill, 1997) was given to formally assess
expressive and receptive language. The Peabody Picture
Vocabulary Test–III (PPVT–III; Dunn & Dunn, 1997) was
administered to obtain standardized receptive vocabulary
scores. Table 1 presents scores for the 3 target children’s
performances on all standardized tests. Although standardized cognitive scores were not available, teachers reported
Evan’s and Tyler’s cognitive skills to be within normal
limits; however, Julie was reported to have a mild developmental delay.
Description of Target Children
Evan. Evan, age 6:1, was described by his teacher as a
child who wanted to play with his peers but who did not
have the social skills needed to join a peer group successfully. Rather than appropriately joining play interactions,
Evan would use more disruptive strategies, such as
Beilinson • Olswang: Facilitating Peer-Group Entry
155
Table 1. Scores for target children on standardized measures.
Participant
Evan
Tyler
Julie
Sex
Age
PPVT–III
TOLD–P:3 (SLQ)
Male
Male
Female
6:1
5:8
6:3
95
72
90
98
84
64
TOLD–P:3 (LiQ)
100
97
79
Note. All names provided are pseudonyms. Age is in years:months. PPVT–III = Peabody Picture Vocabulary
Test–III; TOLD–P:3 (SLQ) = Test of Language Development–Primary:3-Spoken Language Quotient; TOLD–
P:3 (LiQ) = Test of Language Development–Primary:3-Listening Quotient. Standard scores are presented for
the PPVT–III (M = 100, SD = 15) and for the TOLD–P:3 (SLQ and LiQ) (M = 100, SD = 15).
Peer Comparison
Figure 1. (A) Comparison of average number of low-risk entry
behaviors for target children to typically developing peers. (B)
Comparison of average number of high-risk entry behaviors for
target children to typically developing peers.
# of occurrences/10 min
A
40
35
30
25
20
15
10
5
0
Target Children
Evan
40
35
Carter Debbie Annie
Children
Target Children
Comparison Peers
30
25
20
15
10
5
Evan
LANGUAGE, SPEECH,
Julie
0
To further qualify these children for the study, their
peer interaction behaviors were compared to those of 3
children who were enrolled in the program as children
developing typically and were described by teachers as
having age-appropriate social interaction skills (see Figures
1 and 2). These children were selected to represent peerentry performance that was considered typical by the
teachers in the school. Their performance was considered to
be a sample benchmark for comparing the performance of
the target children. The peer interaction behaviors consisted
of the following: (a) peer-group entry behaviors, specifically those identified as low-risk behaviors (e.g., waiting
and hovering, parallel play); (b) high-risk behaviors (e.g.,
making verbal statements); and (c) play behaviors (e.g.,
156
Tyler
Comparison Peers
B
# of occurrences/10 min
knocking over a block structure or taking a toy, in order to
enter the play. Although Evan did not have a medical
diagnosis, he was described as having behaviors similar to
a child with Asperger’s Syndrome (i.e., normal language
skills with deficits in social communication and some
perseverative behaviors). Teachers noted that Evan had
difficulty varying social behavior appropriately, rarely
appeared to have fun, was rarely sought out by peers for
play, rarely suggested new play ideas, rarely smiled at
peers during play, and rarely engaged in play activities
where social interaction might occur.
Tyler. Tyler, age 5:8, had been diagnosed with an
autistic spectrum disorder. He was described by his teachers
as a very shy child who was well-liked by his peers, but
preferred to play alone or with other adults. Teachers noted
that Tyler rarely took turns or played cooperatively with
peers. He had difficulty continuing interactions and using
appropriate social behavior to begin interactions, and was
rarely sought out for play by his peers. Tyler rarely
suggested new play ideas for a play group, rarely shared
materials with peers, and rarely engaged in areas where
social interaction might occur.
Julie. Teachers described Julie, age 6:3, as a very social
child who enjoyed playing with her friends but who had
difficulty appropriately initiating and sustaining her
interactions. Julie was diagnosed as having a global
developmental delay. Teachers noted that Julie rarely
conversed appropriately with peers and rarely suggested
new play ideas for a play group.
AND
HEARING SERVICES
IN
Tyler
Julie
Carter Debbie Annie
Children
solitary play vs. cooperative play). Data were collected on
these comparison peers during one 10-minute observation
period. Figure 1 illustrates the types of entry behaviors
used by the comparison peers during peer-group entry
attempts. These data show that the comparison peers used
fewer low-risk strategies and more high-risk strategies to
initiate peer-group entry than did the target children. Figure
2 illustrates the amount of time that children engaged in
solitary play versus cooperative play. These data indicate
that the comparison peers spent a greater amount of time
engaged in cooperative play than did the target children.
SCHOOLS • Vol. 34 • 154–166 • April 2003
Figure 2. (A) Comparison of average number of solitary play
behaviors for target children to typically developing peers. (B)
Comparison of average number of cooperative play behaviors
for target children to typically developing peers.
A
% of time/10 min
100
Target Children
Comparison Peers
80
60
40
20
0
Evan
Tyler
Julie
Carter Debbie Annie
Children
provider; however, after each child had received 1 week of
treatment, the three classroom teachers shared in the
treatment responsibilities. A peer-entry treatment sequence
was devised based on the research of Dodge et al. (1983),
Hadley and Schuele (1998), and Brinton and Fujiki (2000).
The treatment in this study facilitated the peer-group entry
process by combining the use of a prop with either a lowrisk behavior (mimicking the peer group) or a high-risk
behavior (making a group-oriented statement). Figure 3
displays the five-step treatment sequence used to teach
peer-group entry. The first four steps (“Walk,” “Watch,”
“Get a toy,” “Do the same thing”) were considered the
low-risk strategies and were taught for one session. A fifth
Figure 3. Mayer-Johnson symbols used to teach the peer-group
entry sequence.
Mayer-Johnson Symbol
Peer-Group Entry Teaching Step
B
% of time/10 min
100
Target Children
Comparison Peers
80
60
40
20
0
Evan
Tyler
Julie
Carter Debbie Annie
Children
These data substantiated the teacher judgments that the
target children had deficits in peer interaction behaviors
when compared to peers developing typically.
Procedures
Seventeen children, 9 with disabilities and 8 without,
were enrolled in the full-day kindergarten. All of the
children in the classroom served as host peers during the
study, as the target children were allowed to choose any
classmate with whom to initiate a play attempt. All sessions
occurred during a 45-minute free-choice period four
afternoons per week. During this time, children were
allowed to play with materials anywhere in the classroom.
The classroom staff participated in the intervention either
by directly providing treatment or by supporting a host
peer in an interaction. Staff included a head teacher with a
master’s degree in special education, an assistant teacher
who was completing her degree in special education, and a
full-time aide. The study consisted of two types of sessions: treatment and probe. Treatment sessions were
conducted approximately three times a week for approximately 30 minutes. Probe sessions were conducted once a
week for approximately 10 minutes per child.
Treatment sessions. A second-year master’s student in
speech-language pathology served as the primary treatment
The Picture Communication Symbols ©1981–2003 Mayer-Johnson,
Inc. Used with permission. (www.mayer-johnson.com)
Beilinson • Olswang: Facilitating Peer-Group Entry
157
Teacher training. The primary researcher trained three
teachers during a 11/2-hour training session after the first
child had received one treatment session. The teacher
training consisted of a description of the study, an
overview of peer-group entry research, and an explanation
of the importance of using a prop to facilitate peer-group
entry. The teachers were shown baseline data of the target
children and the typically developing comparison peers to
further illustrate the need for treatment. The treatment
protocol was reviewed in detail, and teachers were free to
ask questions throughout the training. Teachers were
instructed to observe the researcher throughout the
remaining two treatment sessions during the first week of
the treatment phase for the first child in treatment.
Following this observation period, teachers were provided
with the picture symbols and became primary treatment
providers for the first child. They also became primary
treatment providers for the second and third child after
each child had received 1 week of treatment from the
researcher. Although the researcher remained available for
troubleshooting during the free-play session, teachers
remained the primary treatment providers, prompting the
children to use props for peer-group entry 5-8 times
during a 45-minute session. Throughout the treatment
phase, teachers were encouraged to ask questions as
problems arose. All three teachers showed an increase in
their use of props to prompt peer-group entry once they
became primary treatment providers (see Table 3). Toward
the end of the study, teachers’ use of props to prompt
peer-group entry fell to baseline levels; however, this is
likely a result of the increased classroom demands placed
on the teachers by additional end-of-the-year activities and
responsibilities.
Probe sessions. All probe data were collected online by
the researcher or trained coder during a 10-minute observation period during a free-choice activity (see the Appendix
for a sample data collection sheet). Probe sessions were
conducted once a week throughout the study. Probe
step, “Tell an idea,” was added to the sequence to make up
the high-risk strategy. In this step, children were provided
the model of “I have an idea, let’s….” This step encouraged the children to provide some type of verbal statement
about how the prop could be used to enter the peer group.
This high-risk strategy of “telling an idea” was attempted
until treatment was terminated.
Each treatment session began in a quiet hallway where
the peer-entry sequence was reviewed. Mayer-Johnson
picture symbols (King, 1994) were used as a visual aid
when teaching and reviewing the sequence (see Figure 3).
After reviewing the sequence, children were brought back
into the classroom to use their strategies during the freechoice period.
The sequence was taught using a combination of direct
instruction, modeling, and prompting. A sequential treatment protocol based on a hierarchy of least support to most
support was followed (see Table 2). During each 30-minute
treatment session, target children were provided with
approximately eight (range = 5–8) opportunities for peergroup entry. Target children were asked to choose a host
peer and instructed to try to play with him/her. Each target
child was given approximately 2 minutes to attempt peergroup entry and play with the host peer, in the case of a
successful entry. If the target child did not make an entry
attempt, the treatment provider used a prompt, as described
in Table 2, to assist the child. Prompts were provided as
needed until the child successfully made a peer-group entry
attempt. More direct modeling was necessary during initial
weeks of treatment; however, modeling tapered off as
children progressed.
At the end of 2 minutes, the target child and the
treatment provider wrote down the host’s name on a record
sheet. Children’s entry statements also were recorded. The
target child also was offered a sticker or a piece of candy
after each entry attempt. Children were allowed to keep
their earned stickers at the end of each treatment sessions;
they also were offered a treat from a prize box.
Table 2. Sequential teaching hierarchy of support.
158
Level
Prompt
1
No prompt
2
Visual cue
3
Nonverbal prompt (prop)
Researcher (teacher) shows prop to
target child.
“It looks like (target child)
wants to make a cake with
you.”
4
Prop + verbal prompt
“You could use a (prop) to make a
cake just like (host peer).”
“It looks like (target child)
wants to make a cake with
you.”
5
Prop + direct modeling
Researcher (teacher) tells host peer,
“I’m using a (prop) to make a cake
just like (host peer).”
“We’re making cakes just
like you are.”
LANGUAGE, SPEECH,
AND
HEARING SERVICES
Example
Response to host peer
Researcher shows child visual cue that
corresponds with appropriate step in
entry process (e.g., walk, watch, get
a toy, do the same thing, tell an idea).
IN
SCHOOLS • Vol. 34 • 154–166 • April 2003
Table 3. Teacher use of props to prompt peer-group entry as
measured during 10-minute probe sessions.
Teacher
Use of
props before
training
Average use
of props
posttraining
Range of
use of
props posttraining
A
B
C
0
0
0
4.3
3.7
3.3
2–6
0–7
2–5
sessions were designed to be as unobtrusive as possible so
that they might reflect the naturalistic interaction between
teachers and children during the free-choice activity.
Teachers were not given any special instructions during
probe sessions throughout the study.
Coding taxonomy. The taxonomy for data collection
during the probe sessions was based on a coding system
used by Dodge et al. (1983). All coding was done online by
either the researcher or a trained coder. A combination of
interval and event recording was used to code peer-group
entry behaviors (high- and low-risk behaviors), use of a
prop, the target of the entry attempt (peer/teacher), and
play behaviors (solitary or cooperative). For the interval
recording, the 10-minute observation periods were divided
into 10-second intervals to facilitate coding of behaviors.
For the event recording, coders recorded children’s behaviors within each 10-second interval, thus using a partial
interval recording method (see Table 4 for the coding
taxonomy). Probe data consisted of frequency of occurrence
of the following behaviors as they occurred in 10-second
intervals: peer-group entry behaviors (low-risk/high-risk),
prop use during peer-group entry (presence/absence of a
prop), and play behaviors (solitary/cooperative play).
Design. This research employed three case studies. Each
case study used ABA design features to allow for the
investigation of change over time. The ABA features
included (a) baseline, no treatment; (b) treatment using
props and entry behaviors to encourage group entry; and
(c) withdrawal of treatment for examining maintenance. The
ABA features correspond to the addition and removal of
treatment to allow for the observation of behavior change
under different conditions (no treatment and treatment). The
Table 4. Coding taxonomy: Peer-group entry and play behaviors.
Peer-group entry behaviors
Low-risk behaviors
Wait and hover
Target child was within 4 feet of a host peer and available for social interaction
by watching peers who were engaged in some activity.
Parallel play
Target child was playing in proximity to a host peer and with similar materials
in a similar way, but not interacting or sharing a common play scenario.
High-risk behaviors
General statement
Target child made a general statement/question to a host peer that contained
either an object or an action, related to a toy, but not both (e.g., “I have a
block.” “Let’s build.”).
Specific statement
Target child made a specific statement/question containing an object and an
action related to a toy (e.g., “This car can race your car.”).
Use of propa
Target of initiation
Target child used a prop.
a
Target child initiated to a peer or teacher.
Disruption
Target child produced verbal or nonverbal aversive behavior that interrupted or
disrupted play.
Uncodable
Target child produced any behavior that could not be categorized as a play
behavior or an entry attempt.
Play behaviors
Solitary play
Target child was engaged appropriately with materials but was not playing with
other children. The target child could have been engaged with similar materials
to a nearby child or in an area alone.
Cooperative play
Target child was engaged in play with other children around a shared goal/theme
or using shared materials.
No play
Target child was not engaged in any play with materials or people (i.e., during
transition times).
a
Coded when a high-risk behavior was coded.
Beilinson • Olswang: Facilitating Peer-Group Entry
159
second A phase is described as a withdrawal, meaning
treatment is removed. No expectation is held that performance will reverse, but rather this phase allows for
monitoring maintenance (Hayes, Barlow, & Nelson-Gray,
1999). Baselines were gathered over seven sessions to
allow for graduated entry of the target children into the
study. Because this research project did not begin until the
second half of the school year, decisions were based on
time in order to complete the study before the school year
ended. Movement from baseline to treatment phase for the
second child was made after the first child had received
three treatment sessions and one probe session. Movement
from baseline to treatment phases for the third child was
made after the first child had received 2 weeks of treatment. Movement from treatment to maintenance phases was
made after 12 sessions (9 treatment, 3 probe) or until the
target child could perform the treatment sequence independently (with no prompts by a treatment provider) 80% of
the time twice during 1 week of treatment.
Reliability. The primary researcher trained a first-year
speech-language pathology master’s student as the second
coder. Training took place over a 4-week period and
consisted of both repeated dependent and independent
coding sessions in the classroom, followed by discussions
of the coding system. Both the researcher and the coder
observed the target children for five sessions, coding
behaviors together in order to become familiar with the
coding definitions. Once familiar with the coding definitions, the researcher and coder independently coded six
sessions online of various target children. A point-by-point
reliability of 86% was achieved.
Procedural reliability (i.e., treatment fidelity) was
analyzed by having the trained observer record the procedures used by the researcher to teach each child the
treatment sequence. A template was created to guide the
coder in the following categories: introduction of treatment,
teaching the treatment sequence, acting out the treatment
sequence with dolls, reviewing the sequence, and discussing
the procedures for entering the classroom. Reliability
estimates were calculated by dividing the number of
researcher behaviors produced in correspondence to the
treatment plan and multiplying by 100 (Billingsley, White,
& Munson, 1980). One-hundred percent reliability was
achieved for procedural reliability. 1
Measurement reliability between coders was examined
from approximately one third of the data selected from
each phase of the study (i.e., baseline, treatment, withdrawal) for each child for a total of nine sessions. Cohen’s
kappa (Hollenbeck, 1978) was used to determine agreement
between the researcher and the trained coder. “Kappa is
defined simply as the proportion of agreement after chance
agreement is removed from consideration” (Hollenbeck,
1978, p. 91). This statistic was developed for examining
normal scale agreement. A kappa value of 0.86 was
achieved between the researcher and the trained coder. A
kappa value of 0.86 is considered excellent interobserver
1
Procedural reliability for teacher training was not documented because the
training was implemented one time. The exact procedures (as written on
pages 157–158) are a literal transcript of the steps that occurred.
160
LANGUAGE, SPEECH,
AND
HEARING SERVICES
IN
agreement according to guidelines provided by Cicchetti
and Sparrow (1981).
Data analysis. Frequency of occurrence was calculated
for probe data and plotted as follows: peer-group entry
behaviors (low-risk/high-risk), prop use during peer-group
entry (presence/absence of a prop), and play behaviors
(solitary/cooperative play). Data were analyzed by visual
inspection and mean shift between phases using d index
calculations. The d index is used to examine effect size
when the data do not reveal obvious trends (Kromrey &
Foster-Johnson, 1996). The effect-size index (d) provides a
description of the difference in behavior between two
conditions (phases) in terms of standard deviation units.
The d index is obtained by using the following equation: d
= MB – MA/SDA, where MB and MA represent the mean for
treatment phase and baseline phase respectively, and SDA is
the standard deviation of the baseline phase (Kromrey &
Foster-Johnson, 1996). A d index also was calculated for
examining changes between treatment and withdrawal
phases. In addition, teacher exit interviews were conducted.
RESULTS
Peer-Group Entry Behaviors: Examination
of High-Risk and Low-Risk Behaviors
Figure 4 displays the frequency of low-risk (i.e., parallel
play, waiting and hovering) and high-risk behaviors (i.e.,
general and specific statements) used by the children during
the probe session. (The two comparison lines on each graph
represent the average number of low- and high-risk
behaviors respectively used by the peer comparison
children.) Table 5 provides the d index value to statistically
examine performance of each child’s low-risk and high-risk
behaviors between the baseline and treatment phases, and
between the treatment and withdrawal phases. Recall that
the d index provides a measure of change in standard
deviation units when comparing the means in performance
between phases.
In examining Figure 4, one can see that measures taken
during baseline revealed that all 3 treated children used
more low-risk and fewer high-risk behaviors than the peer
comparison group. Further, their high-risk behaviors were
consistently infrequently occurring. During the treatment
phase, all 3 children showed little change in the number of
low-risk behaviors and an increase in the number of highrisk behaviors used during peer-group entry. The d index
value confirms this change, with all 3 children showing an
increase in high-risk behaviors. Evan’s and Tyler’s d index
reflects a change of performance of a little more than 2 SD
on average between the baseline and treatment phases;
Julie’s performance increases on average 41/2 SD between
the baseline and treatment phases. As can be seen in Figure
4 and as reflected by the small change in the d index
(Table 5), all of the children maintained their increased use
of high-risk behaviors during the withdrawal phase. Further,
as Figure 4 reveals, the target children’s performances
appear to resemble those of their comparison peers.
SCHOOLS • Vol. 34 • 154–166 • April 2003
Figure 4. Frequency of children’s use of high-risk behaviors versus use of low-risk behaviors during
peer-group entry attempts during 10-minute probe observation periods. The black circles indicate the
number of high-risk behaviors. The white circles indicate the number of low-risk behaviors. The two
dotted lines indicate average levels of high-risk and low-risk behaviors for the comparison peers.
Table 5. Presentation of d index for low-risk and high-risk data.
d index calculated
between baseline
and treatment
d index calculated
between treatment
and withdrawal
Low-risk data
Evan
Tyler
Julie
–0.324
–1.703
–0.554
0.361
–0.145
–1.139
High-risk data
Evan
Tyler
Julie
2.182
2.333
4.510
–0.124
0.765
–0.107
Peer-Group Entry Behaviors:
Examination of Prop Use
Figure 5 displays the frequency of prop use (presence of
prop) by the children when they were using high-risk
behaviors (i.e., accompanying verbal statements) to enter
peer groups during the probe session. (The two comparison
lines on the graph represent the average number of times
props were and were not used by the peer comparison
group.) All 3 children used props less than their comparison peers did during the baseline phase. Baseline data also
revealed stable, infrequent use of props by the children.
Probe data collected during the treatment phase revealed
that all 3 children showed a slight increase in their use of
props accompanying high-risk behaviors (with the exception
of Julie’s third week of treatment) when compared to
baseline. The children’s prop use rose to levels comparable
with the peer comparisons. The d index data presented in
Table 6 indicate that all 3 children increased their use of
props (presence of props during peer-group entry) on
average during the treatment phase over the baseline phase.
Evan’s use of props increased on average 3 SD; Tyler, 10
SD; and Julie, 21/2 SD. During the withdrawal phase, all
children continued to use props with high-risk behaviors
with little or no change occurring on average between the
treatment and withdrawal phases.
Beilinson • Olswang: Facilitating Peer-Group Entry
161
Figure 5. Frequency of children’s entry attempts with props and without props accompanying highrisk behaviors during 10-minute probe observation periods. The black circles indicate the number of
entry attempts with props. The white circles indicate the number of entry attempts without props.
The two dotted lines indicate the average number of entry attempts with and without props for the
comparison peers.
Table 6. Presentation of d index for absence and presence of
a prop during peer-group entry.
Play Behaviors: Examination of
Solitary and Cooperative Play
Figure 6 displays the frequency of occurrence the
children demonstrated the two play behaviors—solitary play
and cooperative play—during the probe sessions. (Comparison lines on the graph represent the average level of
cooperative play for the peer comparisons.) Data collected
during baseline revealed low levels of cooperative play
across all children. Tyler and Evan demonstrated more
solitary play than cooperative play. All children showed
less cooperative play than the peer comparisons. As can be
seen in Figure 6, all of the children demonstrated an
increase in cooperative play during the treatment phase.
Evan and Tyler showed rather immediate changes in
cooperative play with the introduction of treatment; Julie’s
performance was slower to change. All of the children
demonstrated more cooperative play than solitary play
during the treatment phase. By the end of the treatment
162
LANGUAGE, SPEECH,
AND
HEARING SERVICES
IN
d index calculated
between baseline
and treatment
d index calculated
between treatment
and withdrawal
Absence of a prop data
Evan
Tyler
Julie
–0.622
–0.813
2.350
–0.382
3.186
0.629
Presence of a prop data
Evan
Tyler
Julie
3.080
10.250
2.583
–0.057
0.177
–0.389
phase, all children showed levels of cooperative play
approaching, and in Julie’s case exceeding, the peer
SCHOOLS • Vol. 34 • 154–166 • April 2003
Figure 6. Amount of children’s time spent engaged in cooperative play versus solitary play during
10-minute probe observation periods. The black circles indicate the amount of time spent engaged
in cooperative play. The white circles indicate the amount of time spent engaged in solitary play.
The two dotted lines indicate the average amount of time spent engaged in cooperative play and
solitary play for the comparison peers.
comparisons. The d index values presented in Table 7
reveal Evan’s performance of cooperative play during
treatment to be on average 21/2 SD greater than his performance during baseline. Tyler’s performance is an impressive 13 SD greater during treatment than baseline. Finally,
Julie’s production of cooperative play between the treatment and baseline phases was on average 11/2 SD greater.
Withdrawal data revealed no significant changes in cooperative play compared to treatment for all 3 children,
although as indicated by the d index, Tyler and Julie also
increased their solitary play on average during the withdrawal phase as compared to the treatment phase. All 3
treated children achieved levels of cooperative play that
were more comparable to those of their comparison peers.
Table 7. Presentation of d index for solitary play and cooperative play data.
d index calculated
between baseline
and treatment
d index calculated
between treatment
and withdrawal
Solitary play data
Evan
Tyler
Julie
–0.560
–2.157
–1.063
0.088
3.981
11.500
Cooperative play data
Evan
Tyler
Julie
2.575
13.279
1.545
–0.050
–0.868
0.665
Teacher Data
As discussed in the Methods section (see Table 3),
teachers increased their use of props to prompt target
children to enter peer groups following treatment. Recall
that during baseline, teachers knew the purpose of the
study but did not receive training. During the treatment
phase, teachers received training and observed the researcher for 1 week and then became primary treatment
Beilinson • Olswang: Facilitating Peer-Group Entry
163
providers. Teacher exit interviews also were conducted at
the conclusion of the study. Overall, teachers commented
that the treatment was easy and practical to provide in a
classroom setting. In addition, they noted that they were
able to use the principles of the treatment (i.e., use of a
prop) with children who were not involved in the study.
One teacher remarked that using a prop was especially
helpful for the children with an autism spectrum disorder
because the prop provided a concrete reminder for a child
seeking out a peer for play.
CLINICAL IMPLICATIONS
In summary, these three case studies allowed for the
documentation of change in peer-group entry behaviors
with the introduction of treatment implemented by a
speech-language pathologist and teachers in the classroom.
Of course, given the small number of participants, the
small number of probe sessions, and the relatively small
changes that were observed, conclusions need to be made
with caution. All 3 treated children slightly increased their
high-risk group-entry behaviors during the treatment phase
and continued to use them at a level higher than baseline
through withdrawal. However, all 3 children continued to
use low-risk group-entry behaviors throughout the study.
These behaviors occurred more frequently than they did
for the 3 comparison children. Data collected on the
children’s use of a prop revealed a slight increase in prop
use with the introduction of treatment as well, with a
maintenance of prop use during the withdrawal of treatment. Finally, the children’s play behaviors changed with
the introduction of treatment. All 3 children demonstrated
increases in cooperative play.
The treatment was designed to increase high-risk
behaviors (i.e., general and specific statements to enter peer
groups). Although all 3 treated children demonstrated an
increase in the use of high-risk behaviors following
treatment, 2 of the children continued to use low-risk
behaviors (i.e., parallel play, waiting and hovering). The
continued use of low-risk behaviors indicated that the steps
of waiting and hovering and parallel play appeared to serve
an important role in the peer-group entry sequence.
Although the use of high-risk behaviors is related positively
with increased peer interaction (Dodge et al., 1983; Putallaz
& Gottman, 1981), the use of low-risk behaviors should not
be overlooked. Dodge et al. found that high-risk behaviors
are more likely to be accepted by peers once they have
received positive responses with lower risk behaviors.
Similarly, Craig and Washington (1993) found that children
with SLI use active observation, in the form of waiting and
hovering, as a necessary step in the peer-group entry
process, especially when interactive demands increased.
A distinguishing factor of this treatment was the pairing
of a prop with the use of a high-risk behavior. This
component of the treatment was designed to show the
children how to use language to assist in their peer-group
entry. Following Hadley and Schuele (1998), the researchers attempted to emphasize the contribution of the verbal
164
LANGUAGE, SPEECH,
AND
HEARING SERVICES
IN
component to social interactions. Children were taught to
use a prop to facilitate the formation of an idea or to use
it as a focus of joint attention with the host peer and to
help provide content for verbal statements used for peergroup entry.
Data collected during the treatment phase indicated that
all 3 children showed a slight increase in prop use when
making a statement (i.e., high-risk behavior) to a peer;
however, the prop provided a different type of support for
each child. For Evan and Tyler, the prop seemed to provide
content for language formulation during peer-group entry.
During the treatment phase, both Evan and Tyler were
observed making statements that were related directly to
the prop. For example, Tyler was observed holding a
backhoe and saying, “I have an idea, let’s dig a trench.”
Tyler’s strong change in prop use also may have reflected
his responsiveness to the structure provided by the use of a
concrete object for play ideas. Evan was noted saying,
“Let’s make a fence for the dinosaur,” while holding a
piece of fence.
The use of a prop seemed to provide a different kind of
support for Julie, whose expressive language skills were
more impaired. During the first few weeks of the study,
Julie would get a prop, begin using the prop (i.e., digging),
and say, “I have an idea, let’s do this.” This general
statement provided a source of joint attention for Julie and
the host peer, rather than providing a specific idea for play.
By the end of the treatment phase, Julie began to use more
specific language that related directly to the prop. For
example, she said “Let’s pop bubbles,” while blowing
bubbles with a peer. Perhaps the treatment and prop use
helped facilitate language production for Julie. Although the
exact role played by the prop in facilitating the peer-group
entry process is difficult to define, prop use seemed to be
an important, yet distinctive, part of the process for all 3
children.
Positive changes in the target children were seen in
cooperative play behaviors, as well. Before treatment,
cooperative play levels for all children were well below
cooperative play levels for their comparison peers. Data
collected during the treatment phase revealed increases in
cooperative play levels that were generally maintained
during the withdrawal phase. Further, cooperative play
levels approached, and in Julie’s case surpassed, those of
the comparison peers. No changes occurred in cooperative
play levels until the initiation of treatment for the children.
Again, Tyler demonstrated the greatest amount of change in
his play behaviors in treatment as compared to baseline.
Such remarkable changes may be due to the fact that
although he did not engage in any cooperative play during
baseline, he was afforded multiple cooperative play
opportunities by the treatment package.
Four possible reasons are proposed to explain the increased
amounts of cooperative play resulting from treatment.
• All three target children were required to initiate play
with their peers more than they did on their own
during baseline. By showing interest in playing with
their peers, their peers may have returned the interest
and engaged with them in cooperative play.
SCHOOLS • Vol. 34 • 154–166 • April 2003
• The second explanation for the increase in cooperative
play may have been due to the nature of the high-risk
behaviors. The treatment sequence taught the children
to observe ongoing play with peers and to make a
contribution with a verbal statement, thus making it
more appealing to be invited into the play. Research
has shown that behaviors that make reference to the
group during the peer-group entry process are more
likely to be received positively than those that take
reference away from the group (Dodge et al., 1983;
Putallaz & Gottman, 1981).
• Third, perhaps the host peers began to change their
perceptions of the target children as a result of having
more positive interactions with them over the course
of the study.
• Finally, the teacher training component provided
treatment opportunities throughout the school day,
even when the researcher was not present. Further, the
teachers appeared to become active participants in
implementing the treatment and seemed invested in
facilitating peer-group entry for children who were
exhibiting problems in this aspect of their social
communication.
The results of this study, though preliminary, suggest a
way to teach peer-group entry skills to children with social
communication deficits or language impairments. The data
indicated that this treatment effectively promoted small
changes in social interaction among peers in a relatively
short amount of time (3–5 weeks) with intensive training.
Generalization of findings needs to be guarded; however, the
findings begin to provide some information for treatment
planning. The use of a prop to facilitate the peer-group entry
process appeared to be an effective strategy for these
children when implemented by a speech-language pathologist
and teachers. Having a concrete object proved helpful for
providing content for language formulation and an object for
joint attention. Although more research is needed to explore
this general area, the preliminary data seem promising in
expanding positive peer interactions. Such findings are
encouraging for developing interactions that promote social
interaction of children with language impairments.
ACKNOWLEDGMENTS
This research was completed as a requirement for the primary
author’s master’s thesis. We would like to express our appreciation
to the people who helped to make this research possible. We thank
the teachers and families at the University of Washington Experimental Education Unit for allowing us to be a part of their
classroom. We would also like to thank Allison Eely, who worked
as the secondary coder. We are grateful to Ilene Schwartz, Laura
Sargent, and Truman Coggins for their assistance in this project.
REFERENCES
Billingsley, F., White, O. R., & Munson, R. (1980). Procedural
reliability: A rationale and an example. Behavioral Assessment,
2, 229–241.
Brinton, B., & Fujiki, M. (1997). The ability of children with
specific language impairment to access and participate in an
ongoing interaction. Journal of Speech, Language, and Hearing
Research, 40, 1011–1025.
Brinton, B., & Fujiki, M. (2000, March). Social survival for
children with language impairment. Presentation given at the
Conference for Learning, Seattle, WA.
Cicchetti, D., & Sparrow, S. (1981). Developing criteria for
establishing interrater reliability of specific items: Applications
to assessment of adaptive behavior. American Journal of Mental
Deficiency, 86(2), 127–137.
Craig, H. K. (1993). Social skills of children with specific
language impairment: Peer relationships. Language, Speech, and
Hearing Services in Schools, 24, 206–215.
Craig, H. K., & Washington, J. A. (1993). Access behaviors of
children with specific language impairment. Journal of Speech
and Hearing Research, 36, 322–337.
Dodge, K. A., Schlundt, D. C., Schocken, I., & Delugach, J. D.
(1983). Social competence and children’s sociometric status: The
role of peer-group entry strategies. The Merrill-Palmer Quarterly, 29(3), 309–336.
Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary
Test–Third Edition. Circle Pines, MN: American Guidance
Service.
Guralnick, M. J., Connor, R. T., Hammond, M. A., Gottman, J.
M., & Kinnish, K. (1996). The peer relations of preschool
children with communication disorders. Child Development, 67,
471–489.
Hadley, P. A., & Schuele, C. M. (1998). Facilitating peer
interaction: Socially relevant objectives for preschool language
intervention. American Journal of Speech-Language Pathology,
7, 25–36.
Hayes, S., Barlow. D., & Nelson-Gray, R. (1999). The scientist
practitioner: Research and accountability in the age of managed
care. Boston: Allyn & Bacon.
Hollenbeck, A. R. (1978). Problems of reliability in observational
research. In G. P. Sackett (Ed.), Observing behavior, Vol. II,
Data collection and analysis methods (pp. 79–98). Baltimore:
University Park Press.
King, D. (1994). Boardmaker (Version 1.4) [Computer Software].
Solana Beach, CA: Mayer Johnson.
Koegel, L. K. (2000). Interventions to facilitate communication in
autism. Journal of Autism and Developmental Disorders, 30(5),
383–391.
Kromrey, J., & Foster-Johnson, L. (1996). Determining the
efficacy of intervention: The use of effect sizes for data analysis
in single-subject research. The Journal of Experimental
Education, 65(1), 73–93.
Newcomer, P. L., & Hammill, D. D. (1997). Test of Language
Development–Primary (3rd ed.). Austin, TX: Pro-Ed.
Odom, S. L., & Strain, P. S. (1986). A comparison of peerinitiation and teacher-antecedent interventions for promoting
reciprocal social interaction of autistic preschoolers. Journal of
Applied Behavior Analysis, 19(1), 59–71.
Putallaz, M., & Gottman, J. M. (1981). An interactional model
of children’s entry into peer groups. Child Development, 52,
986–994.
Redmond, S. M., & Rice, M. L. (1998). The socioemotional
behaviors of children with SLI: Social adaptation or social
deviance? Journal of Speech, Language, and Hearing Research,
41, 688–700.
Beilinson • Olswang: Facilitating Peer-Group Entry
165
Redmond, S., & Rice, M. (2002). Stability of behavioral ratings
of children with SLI. Journal of Speech, Language, and Hearing
Research, 45, 190–201.
Received June 27, 2002
Accepted February 3, 2003
DOI:10.1044/0161-1461(2003/013)
Rice, M. L., Sell, M. A., & Hadley, P. A. (1991). Social
interaction of speech and language impaired children. Journal of
Speech and Hearing Research, 34, 1299–1307.
Roth, F. P., & Clark, D. M. (1987). Symbolic play and social
participation abilities of language-impaired and normally
developing children. Journal of Speech and Hearing Disorders,
52, 17–29.
Contact author: Jill Selber Beilinson, 16300 SE 48th Drive,
Bellevue, WA 98006. E-mail: [email protected]
APPENDIX. SAMPLE DATA COLLECTION SHEET USED FOR ONLINE CODING
Play behaviors
1
Solitary
Play (SP)
Cooperative
Play (CP)
Peer entry attempts
Low risk
High risk
Wait & Hover (WH)
Parallel Play (PP)
General Stmt. (GS)
Specific Stmt. (SS)
Play behaviors
Solitary
Play (SP)
Cooperative
Play (CP)
– no prop
(NP)
Peer entry attempts
Prop
Low risk
High risk
Wait & Hover (WH)
Parallel Play (PP)
General Stmt. (GS)
Specific Stmt. (SS)
3
Solitary
Play (SP)
Cooperative
Play (CP)
– no prop
(NP)
Peer entry attempts
Prop
Low risk
High risk
Wait & Hover (WH)
Parallel Play (PP)
General Stmt. (GS)
Specific Stmt. (SS)
Disruption (D)
Uncodable (U)
Not Playing (NP)
166
LANGUAGE, SPEECH,
AND
HEARING SERVICES
IN
+ prop
(P)
Disruption (D)
Uncodable (U)
Not Playing (NP)
Play behaviors
+ prop
(P)
Disruption (D)
Uncodable (U)
Not Playing (NP)
2
Prop
+ prop
(P)
– no prop
(NP)
SCHOOLS • Vol. 34 • 154–166 • April 2003