Available online at www.sciencedirect.

com

Research in Autism Spectrum Disorders 2 (2008) 17–27

http://ees.elsevier.com/RASD/default.asp

Using free operant preference assessments to select toys

for free play between children with autism and siblings
Rachael A. Sautter, Linda A. LeBlanc *, Jill N. Gillett
Western Michigan University, United States
Received 29 January 2007; accepted 7 February 2007

Abstract
Stimulus properties of toys may impact the type and amount of play observed between children with
autism and their playmates. Six children with autism and their siblings participated in an evaluation of toy
characteristics on type of play, problem behavior, social initiations, and responses to social initiations.
Separate free operant preference assessments were conducted with toys in two categories: sensory
stimulating toys and developmentally oriented toys. Highly preferred items and low to moderately preferred
items from each category were then introduced into free play observations with children with autism and
their sibling. Generally, highly preferred sensory stimulating items were associated with more problem
behavior and solitary play while developmentally oriented toys that were moderately preferred produced the
most interactive play and the least problem behavior. Implications for sibling-based play interventions are
discussed.
# 2007 Elsevier Ltd. All rights reserved.

Keywords: Preference assessment; Siblings; Autism; Play

Autism is characterized by pervasive deficits in socialization, communication, and restricted

and repetitive behavior (American Psychiatric Association, 2000). One hallmark communication
impairment is a lack of varied, developmentally appropriate, spontaneous make believe play or
social imitative play. The play of children with autism is often limited and is solitary and
repetitive when it does occur with a tendency towards intense preoccupation with visual
examination of an object or an isolated part of an object (Williams, 2003). Even when matched on
mental age, children with autism exhibit less appropriate play, which is also less varied and
integrated, than non-affected children (Williams, 2003). Failure to develop appropriate play

* Corresponding author at: Department of Psychology, Western Michigan University, 1903W. Michigan Ave,
Kalamazoo, MI 49008, United States. Tel.: +1 269 387 4920; fax: +1 269 387 4550.
E-mail address: Linda.Leblanc@wmich.edu (L.A. LeBlanc).

1750-9467/$ – see front matter # 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.rasd.2007.02.001
18 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27

skills can result in children with autism remaining socially isolated and marginalized (Simpson &
Myles, 1993).
The development of appropriate play skills offers children with autism several benefits. First,
play skills allow children with autism to benefit from play interactions, which are the normative
avenue for obtaining the social, emotional, and cultural experiences needed for typical early
childhood development (Libby, Powell, Messer, & Jordan, 1997). Second, increases in
appropriate play skills are often accompanied by decreases in socially inappropriate behavior
such as self-stimulation and tantrums (Eason, White, & Newsom, 1982; Stahmer & Schreibman,
1992). Third, play skills may improve language gains as suggested by infant studies, which
indicate that turn-taking games and social interactions about objects facilitate early language
acquisition (Bakeman & Adamson, 1984; McArthur & Adamson, 1996; Tomasello & Farrar,
1986; Williams, 2003). Finally, appropriate play skills facilitate integration into mainstream
settings and improve interactions with typical peers, which commonly revolve around toys and
play situations (Restall & MaGill-Evans, 1994).
Behavioral intervention programs have improved play skills of children with autism by
incorporating play activities into social and language instruction (e.g., Charlop-Christy &
Carpenter, 2000; Laski, Charlop, & Schreibman, 1988) and by explicitly targeting functional toy
play and peer interactions with direct intervention (e.g., Harris & Weiss, 1998; Leaf & McEachin,
1999; Maurice, Green, & Foxx, 2001; Maurice, Green, & Luce, 1996). However, curricula and
research studies typically provide little explanation for the selection of specific toys to be used
beyond a brief statement of age and gender appropriateness (Leaf & McEachin, 1999; Lovaas,
2003; Maurice et al., 2001). Some curricula (e.g., Leaf & McEachin, 1999; Maurice et al., 1996)
suggest analyzing the child’s self-stimulatory behavior and incorporating items with similar
properties in play (e.g., toys with lights corresponding to visual self-stimulation), but no
empirical evidence for this recommendation is provided.
A robust literature on preference assessment has illustrated several beneficial procedures for
identifying items that might function effectively as reinforcers when provided contingently on
desired behavior (e.g., Carr, Nicolson, & Higbee, 2000; Cohen-Almeida, Graff, & Ahearn,
2000; DeLeon, & Iwata, 1996; Hanley, Iwata, & Lindberg, 1999; Northup, George, Jones,
Broussard, & Vollmer, 1996; Roane, Vollmer, Ringdahl, & Marcus, 1998). In general, this
literature suggests that highly preferred items are more effective as reinforcers for skill
acquisition than low and moderately preferred items (DeLeon & Iwata, 1996; Fisher et al., 1992;
Northup et al., 1996; Roane et al., 1998). Additionally, presentation of highly preferred items in
a noncontingent arrangement, often referred to as an enriched environment, has proven effective
in decreasing certain types of problem behavior. However, preference assessments have not
been directly used to select toys for play interventions and the finding that high preference
stimulating items produce the best results might not hold true when the target is interactive play
with another child.
A free operant preference assessment is a brief (5 min) assessment involving free access to a
variety of stimuli (Roane et al., 1998). Several items are placed in the environment and the
duration of engagement with each item is recorded as an index of relative preference. Free
operant assessments provide a quick, easy evaluation of preference without removal or
withholding of preferred items or the presentation of specific selection opportunities that might
be perceived as demands. Thus, the likelihood of problem behavior is decreased (Roane et al.,
1998) and the chance of inadvertent social reinforcement that could obscure child preference is
minimized. Free operant assessments could prove useful in identifying toys to include in dyadic
play interventions when the goal is a specific quality or amount of play.
 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27 19

The present study intends to incorporate free operant preference assessments for selecting toys
that could be used in free play between siblings, even when the sibling is not explicitly targeted as
an intervention agent. The study evaluated the utility of a two-part assessment measure for
selecting the most appropriate toys for free play settings for children with autism and a sibling.
The impact of developmental features of toys and preference level on play was examined but no
specific intervention component was included. That is, we assessed whether certain toys, by their
presence alone, resulted in higher frequency or quality of play interactions without any adult
intervention.

1. Method

1.1. Participants and setting

The participants were six dyads, each of which included a boy diagnosed with autism (age 3–9
years) and his typically developing sibling (see Table 1 for a summary of participant
characteristics). All siblings were within 3 years of the age of the target child with autism (age
range18 months to 11 years) and all parents reported poor interactive play between the sibling
dyads. Four of the siblings were male and two were female. Independent private evaluations and
school eligibility determinations established prior diagnoses for the children with autism, but
diagnosis was confirmed using the Gilliam Autism Rating Scale (GARS; Gilliam, 1995). The
GARS is a 56-item rating scale with a reported coefficient alpha of .96 and test–retest reliability
index of r = .88. All parent ratings indicated that the target children had a moderate to high
probability of a diagnosis of autism (see Table 1 for individual scores).
All sessions were conducted at a university in a 7.5 ft  11.5 ft therapy room with a mounted
video camera. The room contained a small table, two chairs, and the toys relevant to the specific
phase of the study. In general, the child with autism and his sibling were alone in the therapy
room while the parents and researchers observed from the next room. However, the siblings in the
two youngest dyads became distressed when the parent left the room so those parents remained in
the open doorway during all sessions but did not interact with the children.

1.2. Materials: toys

The toys used throughout this study were identified through literature reviews on play skills
and interventions and through interviews with school and private service providers. Information
was gathered about the features of toys frequently used in interventions with children with autism
as well as types of toys commonly seen in kindergarten and first grade classrooms. Two different
lists of toys were developed: sensory stimulating toys and developmentally oriented toys. The 12

Table 1
Participant characteristics
Participant Age GARS Sibling Age
Paul 8 98 Sandy 11
Jacob 9 110 Kyle 7
Marty 5 103 Doug 8
Nathan 4 70 Cassie 2
Ben 7 88 Ben 9
Adam 3 75 Eric 18 months
20 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27

sensory stimulating toys contained primary features designed to stimulate a variety of senses
(e.g., blinking lights, spinning parts, noises, music) and were nominated by all providers as
frequently used in intervention and commonly preferred by children with autism. The 15
developmentally oriented toys were relatively low with regard to sensory stimulation and had
primary features that promoted interactive and functional play because they typically required an
interaction with another person to produce the best effect. For example, a bat and ball are
typically enjoyed the most when two people play together.

1.3. Phase 1: preference assessments

Caregivers read the two lists of toys and ranked the top six items from each based on their
knowledge and opinions about the likelihood of their child with autism engaging with the item.
Parents also answered six questions about play between their children and the responses were
used to inform operational definitions for social initiations, responses to initiations, and problem
behavior.
Each child with autism participated in four 5 min free operant preference assessment sessions.
Each session included the six toys identified by the caregiver for one of the categories and two
sessions were conducted with each set of toys. Session order was determined by random selection
for the first and third sessions and counterbalance for the second and fourth sessions. The toys
were equally spaced around the room in a circle on the floor and the child was led to the center of
the circle and shown how to use each toy to ensure contact with the relevant features of the toy.
Following this brief exposure, the child was instructed to ‘‘play with whatever you want.’’ The
researcher remained in the corner of the room and recorded the total duration of the child’s
engagement with each toy using multiple stopwatches. After 5 min, the child was removed from
the room while the toys were exchanged. If a child engaged with only one item during the first
session, that item was not included in the second session for that category to increase the
likelihood of identifying multiple items on a preference gradient. The percentage duration of
engagement for each toy was calculated as the duration of engagement divided by the total
duration of access multiplied by 100 percent.
Based on the percentage duration of engagement in the free operant preference assessment,
four toys were identified: a higher preference developmental toy (Higher/Dev), a lower
preference developmental toy (Low/Dev), a higher preference stimulating toy (Higher/Stim), and
a lower preference stimulating toy (Low/Stim). The higher preference items had the highest
percentage duration of engagement for the relevant category (X = 62 percent; range: 31–100
percent). The lower preference items were preferred over at least one other item in the assessment
and resulted in some level of engagement (X = 11 percent; range: 1–38 percent). These items
were included in subsequent free play sessions in Phase 2.
A second observer independently scored video footage of 25 percent of the free operant
preference assessment sessions across participants and toy type. Total agreement was calculated
by dividing the smaller of the recorded total durations by the larger of the total durations and
multiplying by 100 percent. Agreement percentages were then averaged across all items,
sessions, and participants. Interobserver agreement averaged 99 percent (range 97–100 percent).

1.4. Phase 2: play analysis

Each dyad participated in four 5 min play sessions. Session order was determined by random
selection for each participant prior to the first session. For each session, the two children were led
 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27 21

into the room, which contained one toy identified in Phase 1 (e.g., Higher/Dev, Low/Stim). The
researcher modeled how to interact with the toy, and then instructed the sibling to ‘‘try to get your
brother to play with you, but don’t force him to play’’. After 5 min, the researcher removed the
toy from the room and placed the next toy on the table, modeled interaction with the toy, and
reminded the sibling to try and play with his or her brother.
For the child with autism data were collected on the percentage of intervals with functional play,
the percentage of intervals with inappropriate behavior, and responses to sibling initiations. Four
categories of functional play were coded based on definitions adapted from Baron-Cohen (1987)
and Koegel, Dyer and Bell (1987). Solitary play was defined as using a toy in the intended manner
but more than 2 ft away from another person with out interaction. Parallel play was defined as
playing with a toy appropriately and in a solitary manner but within 2 ft of another person. Pretend
play was defined as using an object as if it were another object (e.g., using a banana as a phone).
Interactive play included any non-disruptive play interactions attempted or achieved between the
child with autism and his sibling. Inappropriate behavior included stereotypical play (i.e.,
repetitive non-functional use of a toy), self-stimulatory behavior (i.e., motor or vocal stereotypies),
and problem behavior (e.g., throwing toys, crying, screaming, hitting).
Responses of the child with autism to sibling initiations were coded as either positive,
negative, or no response. Positive responses included positive verbalizations (e.g., saying ‘‘OK’’,
‘‘let’s play’’, ‘‘more’’) and positive gestures (e.g., leaning or moving towards the sibling or toy,
reaching for the toy, pointing to the toy, joining the sibling in play). Negative responses included
negative vocalizations (e.g., saying ‘‘no’’, ‘‘all done’’, ‘‘no more’’) and negative gestures (e.g.,
moving away from the child or toy, pushing the toy away, attempting to leave the room).
The two dependent measures for the siblings were the percentage of intervals with social
initiations toward the sibling with autism and the percentage of intervals with problem behavior.
An initiation was defined as a verbal or physical attempt to incorporate the other child into play
(i.e. ‘‘let’s play’’, handing a toy, pointing). Problem behavior included throwing toys, crying,
screaming, hitting, kicking, and pushing.
A second observer independently scored a sample of video footage for every child and sibling
dyad. Each 15 s interval was scored as an agreement or disagreement and agreements were
divided by agreements plus disagreements and multiplied by 100 percent to yield the overall
agreement. Agreement percentages were then averaged across participants and sessions for each
relevant dependent measures. IOA for appropriate play behaviors was calculated for 25 percent
of sessions and averaged 90 percent (range 84–95 percent). Agreement for inappropriate
behaviors was calculated for 25 percent of sessions and averaged 96 percent (range 84–100
percent). Agreement for sibling initiations and responses to those initiations was calculated for
21 percent of sessions and averaged 93 percent (range 80–100 percent) for initiations and 94
percent (range 83–100 percent) for responses to the initiations.

2. Results and discussion

The results of the preference assessment sessions for each participant are summarized in
Table 2 indicating that all participants engaged with both types of toys. Five of the six participants
engaged at some level with an equal or greater number of developmental toys than stimulating
toys. No child played exclusively with one educational toy but both Ben and Adam played with
one stimulating toy exclusively in the first sessions. Thus, those items were removed for the
second preference assessment session to ensure identification of a lower preference item that
would result in engagement. Jacob and Adam interacted with only two stimulating toys resulting
22 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27

Table 2
Percent duration of engagement with stimulating toys and developmentally oriented toys during free operant preference
assessment sessions
Participant Stimulating toys Percent duration Developmental toys Percent duration
of engagement of engagement
Paul Light up plastic fan 50 a Legos 46 a
Light up squishy ball 45 Play-doh 46
Cars and track with audio 5b Flash cards 16 b
Light up trucks with audio 0 Animals and barn 0
Jammin draw 0 Markers and paper 0
Trampoline 0 Train set-up 0
Jacob Light up squishy ball 88 a Bat and ball 47 a
Trampoline 4b Train set-up 38 b
Cars and track with audio 0 Play-doh activity set 0
Light up plastic fan 0 Blocks 0
Push button light and sound toy 0 Books 0
Jammin draw 0 Markers and paper 0
Adam Light up trucks with audio 100a Puzzles 33 a
Light up plastic fan 10 b Train set-up 16
Pop up toy 0 Markers and paper 5b
Trampoline 0 Blocks 1
Light up spinning top 0 Books 0
Push button light and sound toy 0 Plane set-up 0
Marty Light up plastic fan 66 a Plane set-up 55 a
Trampoline 32 Books 11
Pop up toy 1b Train set-up 6b
See and say 0 Dinosaur set-up 2
Push button light and sound toy 0 Play-doh activity set 0
Cars and track with audio 0 Plastic food and dishes 0
Nathan Cassette player and tapes 60 a Markers and paper 71 a
Trampoline 48 Play-doh activity set 16
Jammin draw 31 Toy phone 13 b
Light up trucks with audio 13 Bat and ball 1
Light up spinning top 12 b Train set-up 0
Light up squishy ball 11 Puzzle 0
Ben Cars and track with audio 100a Train set-up 31 a
Pop up toy 31 Bat and ball 22
Light up plastic fan 18 b Plane set-up 4b
Push button light and sound toy 0 Play-doh activity set 0
Light up spinning top 0 Legos 0
Light up squishy ball 0 Puzzles 0
Note. This table includes 11/12 stimulating toys (Tickle me Elmo1 was never selected by parents) and 13/15
developmental toys (dolls and stuffed animals were never selected by parents).
a
Items identified as higher preference toys.
b
Items identified as lower preference toys.

in only one possible higher preference item and one possible low preference item for the
subsequent play analysis. Jacob also interacted with only two educational toys, restricting the
possibilities for the subsequent play phase.
In the play analysis, higher preference stimulating toys (Higher/Stim) produced 50 percent
or more engagement for all participants while the higher preference developmental toys
 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27 23

(Higher/Dev) produced 50 percent or more engagement for only two participant dyads (Marty’s
& Nathan’s). Substantial differences in total duration engagement were observed between
stimulating toys with Higher/Stim toys producing an average of 77 percent engagement (range:
50–100 percent) while the Low/Stim toys produced an average of 8 percent engagement (range:
1–18 percent). Smaller differences were observed for the two categories of developmentally
oriented toys with Higher/Dev toys producing an average of 47 percent engagement (range: 31–
71 percent) and Low/Dev toys producing an average of 14 percent engagement (range: 4–38
percent).
Fig. 1 depicts the play analysis for each participant with respect to appropriate play and
nonfunctional behaviors (i.e., self-stimulatory behaviors, stereotypical play, and problem
behavior) of the child with autism. Five of the six participants (83 percent) engaged in interactive
play with a sibling during at least one of the play analysis sessions. Adam engaged in parallel play
only, thus, parallel play is graphed. Four of the five participants with interactive play engaged in
the highest levels of interactive play when the low preference toys were present (Paul, Jacob,

Fig. 1. The percentage of intervals with the most advanced level of observed appropriate play (interactive or parallel) and
the percentage of intervals with inappropriate behaviors for each participant with autism.
24 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27

Marty, and Nathan). During the play analysis phase for Nathan and his sibling, interactive play
only occurred when the low preference developmentally oriented toys were present. Thus, low
preference toys promoted the highest levels of interaction between the participants with autism
and their siblings across the stimulating and educational features of the different toys.
The results in Fig. 1 also show that five of the six participants with autism engaged in
nonfunctional behaviors across sessions during the play analysis phase. Although these results
are varied across preference levels, four of those five participants engaged in higher levels of
nonfunctional behaviors when the stimulating toys were present (see Paul, Adam, Marty, and
Nathan). Thus, the results depicted in Fig. 1 indicate that preference (i.e., low preference toys)
may be an important factor for selecting toys that promote appropriate interactive play between
siblings, and toy type (i.e., stimulating toys) may be important factor to consider for limiting
nonfunctional behaviors that occur during play.
Fig. 2 depicts the percentage of intervals during the play analysis phase that contained
interactions initiated by the sibling (i.e., solid bars) and the percentage of sibling initiations that
were followed by positive responses by the child with autism (i.e., striped bars). There were
sibling initiations and positive responses to these initiations for five of the six participant dyads.
The sixth dyad included the youngest sibling and no initiations occurred. Most siblings initiated
interactions during at least 70 percent of intervals with the highest level of initiations occurring
when the low preference (for the child with autism) toys were present. In addition, all of the
children with autism demonstrated positive responses to the sibling initiations during at least one
play analysis session. Three of the five children with autism responded positively across all four
sessions.
For four of the five dyads, the highest level of positive responses to sibling initiations occurred
when the low preference toys were present. Siblings initiated across toy types, but theirs siblings
with autism were most responsive to initiations with lower preference toys. In general, the results
of the play analysis phase show that low preference toys, independent of toy type, produced the
most sibling initiations as well as the most positive responses to those initiations. Thus, higher
rates and quality of interactions occurred with toys that were preferred by the child with autism
but were not the most highly preferred toys.
Free operant preference assessments were effective at identifying individual toys that
produced differential play patterns for sibling dyads in a quick, fun, and non-demanding manner.
Low preference toys of each type resulted in the highest levels of appropriate play, sibling
initiations, and positive responses to these sibling initiations. Developmentally oriented toys at
both levels of preference resulted in the lowest levels of inappropriate behavior while stimulating
toys were more likely to occasion nonfunctional behavior. The preference assessments and
subsequent play analyses took a total of only 40 min to complete, which is a reasonable time
commitment for most families. However, the variability across participant dyads indicates that
conclusions about overall toy type and level of preferences should not be generalized across
children with autism and their siblings. Rather, an individualized process should be used and can
be conducted quickly.
Although this study identified benefits of using free operant preference assessments to assist in
toy selection for free play opportunities, there are also several limitations that should be noted.
First, only six dyads were included with a substantial range in the age of siblings. The procedure
proved least revealing for the dyad with the sibling under age two. However, additional data
would be necessary before any conclusions could be drawn about appropriate or inappropriate
ages for this type of assessment. Additionally, it would have been beneficial to confirm
preferences that were determined in the previous visit immediately before the play analysis
 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27 25

Fig. 2. The percentage of intervals with sibling play initiations and the percentage of these initiations that were responded
to positively for each participant dyad.

phase. Although there was never more than 1 week between the two phases of this study,
repeating the preference assessment would have ensured the most accurate information when
selecting toys to include in play sessions.
This study provides preliminary evidence for the utility of free operant preference assessment
for purposes other than reinforcer identification, but additional research is needed. Evidence is
also provided to support that notion that preferred, but not highly preferred toys, should be
included in free play situations with children with autism and their siblings. Future studies should
look at the sustained effects of programming toys for free play in natural settings such as homes
or with non-sibling peers in schools. Intervention studies could also examine the effects of
actively intervening with the sibling to ensure high rates of initiations and/or high quality of
appropriate, interactive play efforts. For example, the current study examined a free play
situation but the same toys might reasonably be used in a more structured sibling implemented
26 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27

intervention such as the natural language paradigm (Laski et al., 1988). Future research might
also parametrically examine the effects of toy types and preference level at different ages with
additional participants.

Acknowledgement

The authors wish to thank James Carr and Richard Malott for their helpful comments on a
previous version of the manuscript.

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington,
DC: American Psychiatric Association. Text revision.
Bakeman, R., & Adamson, L. B. (1984). Coordinating attention to people and objects in mother–infant and peer–infant
interaction. Child Development, 55, 1278–1289.
Baron-Cohen, S. (1987). Autism and symbolic play. British Journal of Developmental Psychology, 5, 139–148.
Carr, J. E., Nicolson, A. C., & Higbee, T. S. (2000). Evaluation of a brief multiple-stimulus preference assessment in a
naturalistic context. Journal of Applied Behavior Analysis, 33, 353–357.
Charlop-Christy, M. H., & Carpenter, M. H. (2000). Modified incidental teaching sessions: A procedure for parents to
increase spontaneous speech in their children with autism. Journal of Positive Behavior Interventions, 2, 98–112.
Cohen-Almeida, D., Graff, R. B., & Ahearn, W. H. (2000). A comparison of verbal and tangible stimulus preference
assessments. Journal of Applied Behavior Analysis, 33, 329–334.
DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a multiple-stimulus presentation format for assessing reinforcer
preferences. Journal of Applied Behavior Analysis, 29, 519–533.
Eason, L. J., White, M. J., & Newsom, C. (1982). Generalized reduction of self-stimulatory behavior: An effect of
teaching appropriate play to autistic children. Analysis & Intervention in Developmental Disabilities, 2, 157–169.
Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of two
approaches for identifying reinforcers for persons with severe and profound disabilities. Journal of Applied Behavior
Analysis, 25, 491–498.
Gilliam, J. (1995). Gilliam autism rating scale. Austin, TX: PRO-ED.
Hanley, G. P., Iwata, B. A., & Lindberg, J. S. (1999). Analysis of activity preferences as a function of differential
consequences. Journal of Applied Behavior Analysis, 32, 419–435.
Harris, S. L., & Weiss, M. J. (1998). Right from the start: Behavioral intervention for young children with autism.
Bethesda, MD: Woodbine House.
Koegel, R. L., Dyer, K., & Bell, L. K. (1987). The influence of child-preferred activities on autistic children’s social
behavior. Journal of Applied Behavior Analysis, 20, 243–252.
Laski, K. E., Charlop, M. H., & Schreibman, L. (1988). Training parents to use the natural language paradigm to increase
their autistic children’s speech. Journal of Applied Behavior Analysis, 21, 391–400.
Leaf, R., & McEachin, J. (1999). A work in progress: Behavior management strategies and a curriculum for intensive
behavioral treatment of autism. New York, NY: DRL Books.
Libby, S., Powell, S., Messer, D., & Jordan, R. (1997). Imitation of pretend play acts by children with autism and down
syndrome. Journal of Autism and Developmental Disorders, 27, 365–383.
Lovaas, O. I. (2003). Teaching individuals with developmental delays: Basic intervention techniques. Austin, TX: PRO-
ED.
Maurice, C., Green, G., & Foxx, R. M. (2001). Making a difference: Behavioral intervention for autism. Austin TX: PRO-
ED.
Maurice, C., Green, G., & Luce, S. C. (1996). Behavioral intervention for young children with autism: A manual for
parents and professionals. Austin, TX: PRO-ED.
McArthur, D., & Adamson, L. B. (1996). Joint attention in preverbal children: Autism and developmental language
disorder. Journal of Autism and Developmental Disorders, 26, 481–496.
Northup, J., George, T., Jones, K., Broussard, C., & Vollmer, T. (1996). A comparison of reinforcer assessment methods:
The utility of verbal and pictorial choice procedures. Journal of Applied Behavior Analysis, 29, 201–212.
Restall, G., & MaGill-Evans, J. (1994). Play and preschool children with autism. American Journal of Occupational
Therapy, 48, 113–120.
 R.A. Sautter et al. / Research in Autism Spectrum Disorders 2 (2008) 17–27 27

Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, B. A. (1998). Evaluation of a brief stimulus preference
assessment. Journal of Applied Behavior Analysis, 31, 605–620.
Simpson, R. L., & Myles, B. S. (1993). Successful integration of children and youth with autism in mainstreamed settings.
Focus on Autistic Behavior, 7, 1–13.
Stahmer, A. C., & Schreibman, L. (1992). Teaching children with autism appropriate play in unsupervised environments
using a self-management treatment package. Journal of Applied Behavior Analysis, 25, 447–459.
Tomasello, M., & Farrar, M. J. (1986). Joint attention and early language. Child Development, 57, 1454–1463.
Williams, E. (2003). A comparative review of early forms of object-directed play and parent-infant play in typical infants
and young children with autism. Autism, 7, 361–377.