The League to Fight Neurelitism: Resource Library

A Naturalistic Observation of the Play Behaviour of Children with Autism Spectrum Disorders

Elizabeth A. Holmes



A thesis

submitted in partial fulfillment

of the requirements for the degree

Master of Arts

Brock University


Department of Psychology

BROCK UNIVERSITY

St. Catharines, Ontario

September, 2001
© Elizabeth A. Holmes, 2001


ABSTRACT

Cognitive and social levels of play engaged in by four-to-eight-year-old children
with autism spectrum disorders were examined in naturalistic classroom settings. In
addition, play at home was compared with play at school via parent and educator
interviews. Seventeen school-age children, their educators and caregivers participated in
the study. The most frequently observed play behaviours included parallel-functional
play, adult interactions and solitary-functional play. The play of the children in the
current study was consistent with that of typically-developing preschoolers. Mothers and
educators did not differ significantly in their perspectives of the participants' play
behaviours. In general, educators' and caregivers' reports were positively related to
researcher observations of participants' play behaviours. Methodological considerations
and practical implications for the findings are discussed.


ACKNOWLEDGEMENTS

There are several people to whom I would like to express gratitude for their
assistance on this project. I would like to thank my supervisor, Dr. Teena Willoughby.
Teena taught me the values of motivation, time management and persistence. She has
provided me with a unique perspective on both child and autism research. I am especially
grateful to Dr. Willoughby for helping me with the somewhat daunting task of narrowing
my research focus.

In addition, I would like to offer my thanks to my committee members, Dr. Linda
Rose-Krasnor and Dr. Dorothy Griffiths. Their feedback and insight, based on much
experience in child and youth research, were very helpful to me.

I am grateful to Mr. Paul Perron at the Niagara Catholic District School Board
and Ms. Janet Killins at the District School Board of Niagara for assisting me with
participant recruitment for the study. I would also like to thank the elementary school
principals, educators, parents and participants for allowing me into their schools and
homes to conduct my research.

I owe a vote of thanks to Dr. Michael Ashton, for sharing his statistical expertise
with me. Despite his very busy schedule, Dr. Ashton always made time to listen to my
concerns and provide me with feedback.

I would like to take this opportunity to express my appreciation to Wendy Belbin
for providing me with inter-rater reliability, interview assistance, feedback and emotional
support. In addition, I would like to acknowledge Tammy McKenzie, for being a
sounding board for my ideas, a great statistical resource person and a "reality checker,"


ni


.V ?-;- m


when I've needed one the most. It has been a pleasure to meet all of my colleagues in the
M.A. program at Brock and I have enjoyed the non-competitive, co-operative
environment that we have fostered together.

Most importantly, I would like to thank my family. My parents have selflessly
nurtured me through this experience, and I would not have completed my thesis without
their encouragement. I know I am fortunate to have this kind of support in my life, and it
has often been one of my greatest motivations to succeed.


IV


TABLE OF CONTENTS
INTRODUCTION 1

Defining Play 3

Theoretical Perspectives 3

Classical theories 3

Psychoanalytic approach 3

Constructivist approach 4

Sociocultural approach 5

Atheoretical Perspectives 6

Types of Play Behaviour 7

Cognitive Levels of Play 7

Sensorimotor/Exploratory play 7

Functional play 7

Pretend play 7

Dramatic play 8

Constructive play 8

Games-with-rules 8

Social Levels of Play 9

Solitary play 9

Parallel play 9

Group play 10

Unoccupied/Onlooker 10

How Play Benefits the Child 10

Constructive Playthings 11

Fluid constructive toys 11

Structured constructive toys 12

Functional Playthings 13

Dramatic Playthings 14

Symbolic Playthings 14

Social Play 15

Play Behaviour in Children Whose Development is Not Typical 16

Autism 17

Diagnostic criteria 17

General play behaviours 19

Cognitive levels of play 22

Social levels of play 24

Problems with current research 25

Summary 28


Rationale 28

METHOD 30

Participants 30

Procedure 34

Measures 37

Background information 37

Verbal mental age 37

Play information provided by parents and teachers 38

Play scale 42

RESULTS 48

Play Activities Engaged in by Caregivers with Participants 48

Research Question One: The Cognitive and Social Levels of Play Engaged in

by Children with Autism 49

Observed Cognitive and Social Behaviours 49

Research Questions Two and Three: Caregiver and Educator Perspectives on

Participant Play Behaviours 51

Comparing Parent and Educator Perspectives on Participant Play

Behaviours 52

PPBS 52

Interview 56

Research Question Four: Comparing Observations with Educator and

Caregiver Reports 64

Observed vs. Reported Behaviours 65

Comparing mean responses to the PPBS and the POS 65

Correlating observed and reported behaviours 67

DISCUSSION 68

Observed Play Behaviours 68

Relationship to typically-developing children of the same age. . . .73

Parent and Educator Perspectives 74

vi


Comparing Parent and Educator Perspectives 75

Comparing Adult Perspectives with Observed Behaviours 76

Methodological Considerations 76

A Comparison Between Maternal and Teacher Reports And

Observed Play Behaviours 76

General Methodological Considerations 77

Summary 78

Suggestions for Future Investigation 80

Conclusions and Practical Implications 82

REFERENCES 84


Vll


LIST OF TABLES
TABLES PAGE

1 . Sex, Diagnosis, Chronological Age and Verbal Mental Age (as Assessed by
PPVT-R) of Participants 31

2. Variables Measured by the POS and their Operationalizations .44

3. Percentages of Total Observed Play Behaviours Across all Children

in Descending Rank 50

4. Percentages of Total Observed Play Behaviours Represented by each Social and
Cognitive Category in Descending Rank 52

5. Summary of Paired-samples T-tests, Means and Standard Deviations for Mother and
Educator Perspectives of Participant Play Behaviours (PPBS) 54

6. Summary of Paired- samples T-tests, Means and Standard Deviations for Averaged
Mother and Educator Perspectives of Participant Play Behaviours (PPBS) 55

7. Summary of Paired-samples T-tests for Comparisons Between PPBS
Subcategories 56

8. Summary of Paired-samples T-tests, Means and Standard Deviations for Mother and
Educator Perspectives of Participant Play Behaviours (Interview Question 1) 58

9. Summary of Paired-samples T-tests, Means and Standard Deviations for Mother and
Educator Perspectives of Participant Play Behaviours (Interview Questions

3, 5,7 & 8) 60

10. Mothers' and Educators' Scores on the PPBS, as Compared with Observed
Behaviours 66

1 1 . Correlations Between Observed and Reported Play Behaviours 68


Vlll


LIST OF APPENDIXES
APPENDIX PAGE

A Recruitment Materials

1 . Informed Parental Consent Letter 92

2. Informed Educator Consent Letter 94

B Parent and Educator Forms and Questionnaires

1 . Parental Consent Form 97

2. Educator Consent Form 99

3. Participant Background Information 101

4. Preschool Play Behaviour Scale 104

5. Parent/Teacher Interview Protocol 108

6. Play Observation Scale 113

C Debriefing and Feedback Forms

1 . Parental Debriefing Form 114

2. Educator Debriefing Form 115

3. Feedback Letter 116

D Regrouping of PPBS items 117


IX


INTRODUCTION

Theoretically, play has been linked to many areas of development, including
intellectual, social and emotional growth. Play promotes intellectual development by
providing the young person with a context in which he or she can practice his or her
language and communication skills. Additionally, the child is granted the opportunity to
employ his or her mind to fantasize, to plan strategies, and to solve problems (Hughes,
1999). Hill and McCune-Nicholich (1981) detected a positive relationship between level
of symboHc play engaged in (i.e., presymbolic scheme vs. single pretend acts vs.
symbolic play sequences vs. planned pretend) and mental age in their observations of
children with Down Syndrome playing with a primary caregiver. The correlation was
stronger for mental age with symbolic play than for chronological age with symbolic
play. Arguably, increasingly sophisticated levels of symbolic play are indicative of
intellectual growth (Hill & McCune-Nicholich, 1981).

Social development also occurs through play. The child acts out different roles
and hence learns how to take the perspectives of others. In addition, role-playing prepares
the child for adult life and the many complex roles an adult plays. In order to play
successfully, children must interact with each other to coordinate activities and resolve
disputes (Hughes, 1999). Interpersonal relationships in childhood also advance social
development by teaching the child mutual respect for the needs of others, equality, and
reciprocity as a tool to get one's own needs met (Sullivan, 1953). In their research on
play in children with Down Syndrome, Hill and McCune-Nicholich (1981) noted a
positive correlation between level of symbolic play achieved and social responsiveness to


the participant's mother and the researcher. Although the authors note this may be due to
the relationship between cognitive and affective development (i.e., that higher levels of
cognitive development are related to more sophisticated social behaviour, regardless of
the level of social play achieved), they don't rule out the possibility that increased
complexity in symbolic play also contributes to cognitive development (Hill & McCune-
Nicholich, 1981). In other words, symbolic play may contribute to both cognitive and
social development, as well as there being a significant positive correlation between
cognitive and social development.

Finally, emotional development is advanced through play. Conflicts and
complaints will arise during play and the child must learn to express what he or she is
feeling, as well as how to compromise with his or her playmates (Hughes, 1999). It is
easier to resolve disputes when the child is able to take the perspective of others and thus
to negotiate effectively with them (Rubin & Coplan, 1998).

Arguably, then, a child who does not get the opportunity to play at increasingly
sophisticated cognitive levels may miss some important developmental milestones.
Young people who have developmental delays, in addition to not having many
occasions to play, are at a double disadvantage. For instance, children with autism
experience difficulties with social interactions and ofl:en do not play with their peers
(Lord, 1984; Tilton & Ottinger; Winzer, 1987). If the child with autism is not receiving as
many chances to play as his or her typically-developing peers, he or she might be missing
the chance to develop due to the autism as well as fewer play opportunities.

In this thesis, I will discuss the quality of play behaviour exhibited by children


with autism. I will demonstrate the need for more research on what children with autism
do during their free play time at school and in the home. I will attempt to meet this need
by conducting a naturalistic observation of their play in free play situations.

Defining Play
Theoretical Perspectives

Classical theories . Among classical theorists, four ideas existed as to why humans
play (see review by Rubin, Fein, & Vandenberg, 1983). The main tenet of the surplus
energy theory was that immature forms of life (e.g., infants and children) are not
responsible for their own survival. For young people, the energy that would have gone
into maintaining oneself is pent up and must be spent somehow. Hence, they play.
Conversely, relaxation and recreation theorists held that play is engaged in to restore
energy lost in day-to-day activity. These theorists believed that play has restorative
properties. The third idea, practice theory, is based on the notion that play is practice for
adulthood. Species that are more complex enjoy longer childhoods, which give their
young more of an opportunity to develop skills used when they mature. Finally, the major
thrust of recapitulation theory is that the cognitive developmental stages of play reflect
human evolution. For instance, play begins with banging objects and gradually progresses
to more sophisticated activities, such as role-playing. Similarly,

humans began primitively (e.g., hunting together) and advanced to contemporary social
interactions (e.g., business meetings).

Psychoanalvtic approach . When examining the extant literature on play
behaviour, one is sure to encounter various definitions of the term "play." Freud (1959)


viewed play as an opportunity for children to achieve wish fulfillment and mastery over
past traumatic events. According to the psychoanalytic perspective, children use play to
project their inner thoughts and feelings (Wolfberg, 1999). It can be an opportunity for
the child to vent socially unacceptable emotions and impulses in a safe context (Rubin et
al., 1983). Play is therapeutic because the child can deal with inner conflicts and anxieties
as well as reduce drives. Trauma can be mastered by repeating the components of a
difficult event during play. This allows children power over events during which they had
previously been powerless (Rubin et al., 1983).

Constructivist approach . Piaget (1962) saw play as a state of imbalance between
intellectual assimilation and accommodation. Assimilation is defined as the incorporation
of events, objects or situations into existing ways of thinking (Rubin et al., 1983).
Accommodation is the modification of internal schemes to fit a changing cognizance of
reality (Reber, 1985). In play, assimilation overcomes accommodation to suit the child's
purposes. For example, accommodation might occur when the child imitates a behaviour.
Assimilation would overcome accommodation when the child is able to play
spontaneously while including a behaviour that had earlier required imitation (Rogers,
1988).

Piaget believed that play development occurs within the individual child
(Wolfberg, 1999). Children are intrinsically motivated to play because they begin to
master their play objects and derive satisfaction from their newly sophisticated ways of
thinking and behaving. Higher-level cognitive abilities emerge as a result of the object
and role manipulations inherent in play (Wolfberg, 1999). Piaget maintained that play


occurs for its own sake. It is the repetition of means that might have been connected to an
end previously, but the focus shifts to the means. Eventually, this constant repetition
leads to the mastery of a skill (Rubin et al., 1983).

Not unlike Freud, Piaget made reference to the possibility that children might
change the meaning of distressing events through play (Rubin et al, 1983). In addition,
Freud and Piaget both believed that play begins to disappear with the arrival of rational
thought. Freud believed this was because the ego had developed to the point where it
would no longer allow the id to express itself as freely (Rubin et al., 1983).

Sociocultural approach . Like Freud, Vygotsky (1967) believed that play was the
resuh of the child's unfulfilled desires. To achieve these desires, the child enters his or her
imaginary world. Although Freud and Vygotsky both viewed play as a release from
tension, Vygotsky saw play as relief from general, day-to-day tension, whereas Freud
saw it as occurring in reaction to specific events (Rubin et al., 1983).

Both Piaget and Vygotsky perceived play as an activity engaged in for
the satisfaction it brings. Satisfaction is derived from the activity itself as well as the
mastery of the skills involved in play (Wolfberg, 1999).

Unlike Piaget, Vygotsky viewed play as a shared experience, in which children
interpret play situations in the context of the values and knowledge upheld by their
culture (Wolfberg, 1999). The play produced by children represents what they believe
their society values. Vygotsky claimed that the transmission of cultural knowledge
through social interaction was necessary to the development of the mind. Hence,
according to Vygotsky, play is influential to human development (Wolfberg, 1999).


O' /,•


Atheoretical Perspectives

Reber (1985) described play as a diversion or recreation, an activity not
necessarily taken seriously. He also stated that one abridged dictionary he consulted gave
55 different meanings for the term.

Rubin and his colleagues (1983) offered one of the most comprehensive and
concise definitions of play. This operationalization will be used as a working definition of
play for the purposes of the proposed study. According to these researchers, play is
comprised of five major factors. First, play is intrinsically motivated. This means that it is
engaged in for pure pleasure, not to satisfy the needs of others. This factor mirrors the
beliefs of Vygotsky and Piaget. Second, the individual freely chooses the activity viewed
as play. If a child is pressured into an activity, he or she may see it as work rather than
play. Third, the pastime must be pleasurable. If the activity causes the child to feel stress
or tension, it cannot be considered play. Vygotsky and Piaget also mentioned that play is
pleasurable. Fourth, play can be nonliteral. The player is given the opportunity to use
make-believe to twist reality to fit his or her needs. Finally, the child actively engages in
play. This means that play requires complete physical, emotional and psychological
involvement. If a child is passively observing what is going on, he or she is not playing.
The second, fourth and fifth factors were not mentioned by Freud, Piaget or Vygotsky,
however, they do not directly contradict them. Despite the fact that a clear definition of
"play" has not yet been established, research continues to be conducted on many different
aspects of play behaviour.


Types of Play Behaviour
Cognitive Levels of Play

Sensorimotor/Exploratory play . In addition to the many different meanings of the
term "play," there are also several distinct cognitive categories of play. The typically-
developing infant participates in sensorimotor play, operationalized as the extent to which
the child acts on objects without taking the fiinctional features of the objects into account
(Libby, Powell, Messer, & Jordan, 1998). Banging a rattle on the floor or exploring it
with one's mouth are examples of sensorimotor play. The child is not shaking the rattle,
which is the appropriate way to play with it. Rather he or she is exploring its physical
properties.

Functional play . Eventually, typically-developing young children begin to take
the functional features of objects into account. Functional play, as the name implies,
involves using an object as its function denotes (Libby et al., 1998). Children engaging in
functional play will often repeat their behaviours over and over again, seemingly for the
pleasure the behaviours bring them (Rubin & Coplan, 1998). For instance, at this
cognitive level, the child would shake the rattle, rather than mouth it or bang it, because it
is the appropriate thing to do with it.

Pretend play . Older children (i.e., approximately 3 to 6 years of age) exhibit play
behaviour that is not dependent on the presence of a physical object
(decontextualizations) and increased use of items to represent absent objects (object
substitutions, Rubin et al., 1983). Hence, play is symbolic in nature. According to
Malone and Stoneman (1990), pretend play involves an element of pretense, or "make-


8


believe." For the child, objects represent whatever is needed to aid in maintaining the
pretense. Symbolic play consists of three principles. First, the child must engage in object
substitution, using an object as if it were something else. Second, the child must attribute
false properties to an object by believing an object possesses qualities that it does not
have. Third, the child must refer to an absent object as if it were present (Libby et al.,
1998).

Dramatic play . Dramatic endeavors take up more of the typically-developing
child's playtime as he or she reaches approximately 6 or 7 years of age (Rubin et al.,
1983). During dramatic play, the child creates an imaginary situation and may engage in
role-taking (McCabe, Jenkins, Mills & Dale, 1996). Turn taking and role-playing take on
a ritualistic quality as children adhere to the rules they have created (Rubin et al., 1983).

Constructive play . In addition to sociodramatic play, other new forms of
behaviour evolve with development. Constructive play occurs when the child uses objects
to build or create something (McCabe et al, 1996). The child's goal in constructive play
is to build something that will remain after play is finished (Rubin & Coplan, 1998). For
instance, he or she might build a tower from blocks or a castle from sand.

Games-with-rules . Games-with-rules take place when the players agree on rules,
roles to be enacted and the division of labour before the games commence. This activity
is relatively rare in preschool and kindergarten-aged children, and becomes more
prevalent in older children. Games-with-rules are often engaged in by adults as well
(Rubin & Coplan, 1998).


Social Levels of Play

Solitary play . Typically-developing infants and toddlers generally engage in
solitary pretense play (Rubin et al., 1983). Rubin and Coplan (1998) operationalized it as
play, taking place apart from other children, at a distance of greater than three feet. These
children have not yet developed an awareness of others. Thus, they play alone.
Alternatively, older children might occasionally choose to play alone, with their backs to
others. They may even be playing with a different type of toy than their peers.

Parallel play . Eventually, they become aware of others, however they do not
know how to ask them to play with them. Therefore, they play near them. This behaviour
is known as "parallel play" because the child does not make an effort to organize
activities with a playmate (Reber, 1985). Rather, he or she will play independently even
when in close proximity to a peer. Rubin and Coplan (1998) define parallel play as
independent play occurring within three feet of other children. Parallel play is the most
common form of social interaction in young children (Rubin & Coplan, 1998).

Although previous research suggests that solitary play is the least mature form of
social play (e.g., Parten, 1932), Rubin, Maioni and Hornung (1976) believe that this is not
the case. These researchers suggest that parallel play actually represents the lowest level
of the social-cognitive play hierarchy. Children who play in a solitary fashion may
choose to do so as an escape, whereas children who engage in parallel play may do so
because they lack the ability to take the perspective of others or to initiate group play
(Rubin et al., 1976). Alternatively, children might prefer one form of play over the other
due to their personalities. Perhaps a shy child enjoys parallel play because it is not


10

necessary to initiate anything with a peer, yet peers are still nearby. Perhaps a withdrawn

child prefers solitary play because it does not require social interactions and it is

enjoyable.

Group play . Young people may also participate in group play, defined as playing with

other people with a common purpose to the activity.

Group play was originally split into two categories, associative play and
cooperative play (Parten, 1932). Associative play was defined as playing with other
children without role assignments or other organization to the activity. Cooperative play
differed from associative play in that it was group play in an organized fashion (Parten,
1932). However, Rubin, Watson and Jambor (1978) collapsed these two categories into
one, "group play." This was done because "associative play," and "cooperative play,"
appeared to be used interchangeably when coding identical activities in preschool and
kindergarten children (Rubin et al., 1978). As a result, cooperative and associative play
will not be used as separate categories in this thesis.

Unoccupied/Onlooker . Rubin and Coplan (1998) also made reference to two other

types of social participation, unoccupied behaviour and onlooker behaviour. Unoccupied

1
children lack focus in their activities. Usually, they will either stare into space or wander

about aimlessly. Those who choose to play the role of onlooker will watch the play of

others, but will not attempt to involve themselves with the activity (Rubin & Coplan,

1998).

How Play Benefits the Child

According to Hughes (1999), virtually no experimental research on how specific


11

types of play affect the child's cognitive development has been done. Most of the extant
research is correlational in nature, meaning that conclusions about causality cannot be
inferred from it. In addition, confounds such as the child's physical and social
environment must be considered. For example, a child who is trying to play in a noisy
room may find it difficult to concentrate or focus. Also, a child who is neglected by his or
her parents and peers may not find play as enjoyable as a child who is securely attached
and popular at school.

Nonetheless, there is some indication that parental involvement and the
availability of a variety of toys in early childhood is related to intellectual development
later in life (Barnard, Bee, & Hamilton, 1984; Bradley, 1986; Bradley & Caldwell, 1984).
The benefits gained fi-om play with specific types of toys and non-toys will be outlined
below.
Constructive Plavthings

Play with constructive toys influences perceptual performance in the form
of block building, puzzle forming, and right-left orientation. According to Wolfgang and
Stakenas (1985), there are two types of constructive playthings, fluid constructive toys
and structured constructive toys.

Fluid constructive toys . Fluid constructive toys have a flexible quality to them.
Their physical properties change when they are played with. Examples include paint or
clay. These toys are used to create unspecified products.

The benefits of play with clay to the intellectual development of the child include
an understanding of the Piagetian concept of conservation (Rubin et al., 1983).


12


Specifically, the child discovers that the amount of clay remains the same if its physical
appearance is altered, as long as nothing is added to or subtracted from it. This discovery
can then be generalized to other substances.

Water is another example of a fluid constructive toy. By handling toys in the
water, the child will learn whether or not an object can float and what characteristics are
needed for an object to float. When they have access to soap, children can learn to create
bubbles of different sizes. Measurement is enhanced through water play as the child
learns the answers to such questions as how many times does the small jar have to be
filled and emptied into the large jar before the large jar is filled? Like clay, water aids in
teaching the child the concept of conservation. The child learns that the amount of water
remains the same, despite different sizes and numbers of containers, provided no water is
added or taken away (Hughes, 1999). The concept of reversibility is also learned through
water play (Hughes, 1999). The child discovers that emptying a large container of water
by filling a smaller container several times can then be reversed (i.e., pour the water back
into the large container by filling the smaller container the same number of times again).

Structured constructive toys . The physical properties of structured constructive
toys don't change when they are played with. Examples include Lego, blocks and
puzzles. Structured constructive toys are used to create a specified product. They are
believed to contribute to verbal, perceptual, quantitative and memory development
(Wolfgang & Stakenas, 1985).

Specific structured constructive toys benefit children's development in distinct
ways. For example, blocks aid the child in understanding the concept of measurement.


13

Questions such as how many blocks does it take to build a roadway from one end of the
block center to the other frequently come up when children play with blocks. They learn
how to answer these kinds of questions on their own (Hughes, 1999). In addition, blocks
have the potential to teach the child about equivalency, the understanding that any area
can be divided into a number of different- sized individual pieces and that a certain
number of pieces of one size makes up one piece of another size. For example, if two
children wish to build the exact same tower from blocks, and only enough 12-inch blocks
are available to build one structure, the other child would have to figure out how many 6-
inch blocks it would take to build the same tower (Hughes, 1999). Also, children who
play with blocks learn how to balance them (i.e., heavier blocks on the bottom, lighter
blocks on the top, otherwise the structure will topple). A better understanding of space is
fostered through block play as the child builds upwards, outwards, horizontally and
vertically (Hughes, 1999). Even the experience of putting blocks away may be
educational to the child, as blocks are usually put away according to their size and shape.
This develops a skill known as logical classification (Hughes, 1999).
Functional Playthings

Wolfgang and Stakenas (1985) emphasized the importance of play with
microsymbolic toys to the intellectual development of children. Microsymbolic toys are
miniature versions of real items. For example, toy cars are often quite detailed and
include many of the accessories that real cars have. Many have trunks, doors that open
and close, and tiny steering wheels. Functional playthings are believed to influence
memory development (Wolfgang & Stakenas, 1985). For instance, the child who plays


> :iii;j ;;>V


14


with the tiny car will remember that his or her parents' car has doors, a steering wheel
and a trunk and that these accessories perform specific functions.
Dramatic Playthings

Macrosymbolic toys are child-size playthings, such as the props used in dramatic
play. Macrosymoblic play is believed to be associated with memory, perceptual
performance and quantitative skills (Wolfgang & Stakenas, 1985). A connection between
predisposition to play in an imaginative fashion and the ability to solve problems
creatively has also been found (Rubin et al., 1983). Decentration, the ability to recognize
that an object can serve more than one purpose, is also evidenced in dramatic play. For
instance, a child with decentering skills recognizes that the plastic laundry hamper
serving as a cage for his or her stuffed animal is also a plastic laundry hamper and maybe
it was a boat for crossing the Nile a few minutes earlier (Hughes, 1999).
Symbolic Playthings

Hughes (1999) proposed a link between symbolic play and language
development, in that both require the ability to mentally represent one's surroundings to
oneself In addition, he suggested that those children who are impaired in symbolic play
also have deficits in their language development. For instance, it is possible that children
who are asked to act out a story they just had read to them will have a better chance of
understanding what they heard, as well as seeing the situation fi-om the perspective of the
characters whose roles they are playing, than those children who simply hear the story
(Saltz, Dixon, & Johnson, 1977; Silvern, WiUiamson, & Waters, 1982).


15


Social Play

In addition to the many intellectual benefits to the child's development listed
above, play aids the child in socialization. First, social play helps the child to learn the
rules underlying both play and social interactions in general (Hughes, 1999). For
instance, through social play, the child learns that turn-taking is an important element of
interacting socially, which must be respected. Second, social integration can be
encouraged with the right playthings (Hughes, 1999). For example, blocks are an
appropriate toy for children who usually have difficulties integrating socially. With
blocks, children can choose the level of social involvement that they are comfortable
with. Blocks can be played with individually, in dyads, or in groups. In fact, children who
begin block play in a solitary fashion will often gradually become socially involved with
other children through the blocks (Rogers, 1985). The same effect can sometimes be
observed when clay is the common plaything or during music time (Hartley, Frank, &
Goldenson, 1952). Although music time is not technically free play time, the same effect
might be observed when a group of children decide to play a game involving music, such
as clapping games, running and chasing games using rhyme and skipping songs. Third,
social play affords the child the opportunity to express his or her inner thoughts and
feelings. For instance, creative movement exercises during class time encourage the child
to enact an emotion through facial expressions and body conduct (Werner & Burton,
1979). Children might choose to repeat these exercises during free play time. Fourth,
social play aids the child in comprehending the perspectives of others in various social
situations (Hartley et al., 1952). For example, in sociodramatic play, the child takes on


16


different roles and learns to see the world through the eyes of other people. Role taking
also aids the child in making strong friendships as the child must be able to understand
how his or her actions will affect others (Harter, 1983). Fifth, social play encourages
cooperation among the players. Group sociodramatic play requires organization and
adherence to rules (Hughes, 1999).

De Kylen and Odom (1989) suggested that social play might be enhanced
(particularly in children with disabilities) in situations where teachers structure an activity
a great deal. These findings occurred independently of teachers interacting with their
students. In fact, children interacted more with their peers than with their teachers in
highly structured activities (De Klyen & Odom, 1989).

Those children who are competent players enjoy all of the developmental benefits
of play discussed above. What about those who do not play in a sophisticated fashion?
Do they miss the social, emotional and intellectual developmental opportunities
gained by their peers? Should this indeed be the case, it would be important to encourage
play in children lacking the ability or motivation to do so independently. Next, I will
discuss the difficulties experienced by children whose development is not typical
(specifically those with autism) when they are engaged in play behaviour.

Play Behaviour in Children Whose Development is Not Typical

Typically-developing children should be able to master all of the cognitive and
social forms of play listed above by approximately age seven years (Rubin et al, 1983).
However, this play progression may not be true of children whose development is not
typical. One type of developmental delay that has been addressed in the play literature is


1 . .,


17


autism.
Autism

Diagnostic criteria . Autism is classified as "Autistic Disorder" in the Diagnostic
and Statistical Manual of Mental Disorders (4^^ Ed.) under the category of Pervasive
Developmental Disorder (American Psychiatric Association, 1994). Rett's Disorder,
Childhood Disintegrative Disorder, Asperger's Disorder and Pervasive Developmental
Disorder-Not Otherwise Specified are also included in this category. According to Tager-
Flusberg, Joseph and Folstein (2001), autism is now considered a spectrum of disorders
consisting of repetitive behaviours, deficits in social interaction and language-related
difficulties. Disorders within the spectrum include autism with an additional medical
condition, Kanner-type autism (the classic definition), PDD Not otherwise specified
(autism symptoms that do not quite meet the diagnostic criteria) and Asperger's
Syndrome (normal language development, but social deficits consistent with autism).
Some controversy exists as to whether or not high-functioning autism can be
distinguished from Asperger's Syndrome.

Autism is considered a lifelong condition that affects twenty in ten thousand
children (when including autism of both Kanner's and Asperger's definition) across all
racial and socio-economic groups (Wolfberg, 1999). Boys with autism outnumber girls
with autism by a roughly 2-4: 1 ratio. In an estimated sixty to eighty per cent of cases of
autism, mental retardation is present. However, children with autism can possess average
and above intelligence as well (American Psychiatric Association, 1994).

Symptoms include qualitative impairments in social interaction, as characterized


rx


18

by a lack of response or an unusual response to external events or actions (American
Psychiatric Association, 1994; Winzer, 1987). For example, social withdrawal, self-
injurious, aggressive and asocial behaviour might sometimes occur. People with autism
also display qualitative impairments in communication (e.g., problems with speech and
language including mutism, echolalia and perseverative or repetitive speech). The third
most prominent symptom is restricted, repetitive and stereotyped patterns of behaviour,
interest, and activities. For example, a child with autism might engage in repetitive motor
activity (such as hand flapping), constant repetition of pleasurable activities (such as
watching the same video several times), or an almost obsessive interest in a topic (such as
vans, Tager-Flusberg et al., 2001; Tilton & Ottinger, 1964). An individual with autism
would appear very concerned with sameness or constancy of order (Geneva Centre, 1997;
Tilton & Ottinger, 1964). Although these are general symptoms of autism, they manifest
themselves in unique ways. No two individuals with autism are going to exhibit the exact
same behaviours (Wolfberg, 1999). Symptoms of autism may be lessened greatly by
providing intervention as early as possible (i.e., at the preschool level). The need for early
intervention has become more recognized in recent years (Maurice, 1993; Rutter, 1973).
Several theories as to the cause of autism have been offered (Wolfberg, 1999). It
was originally believed that autism was the result of being raised by an emotionally
detached mother (Baron-Cohen & Bolton, 1993; Bettelheim, 1967). In recent years,
however, it has been suggested that autism has biological roots. Present research indicates
that brain anomaUes in children with autism have resorted from such contributory agents
as viral infections, complications during delivery, metabolic abnormalities, genetic


r.5


,' i-


19

factors and possibly environmental toxins (Baron-Cohen & Bolton, 1993). The general
consensus in current research is that autism has a genetic predisposition. However, it is
possible that this is offset by environmental or biological factors (Aylott, 2000). Thus,
there are several different ways in which an individual child might develop an autism
spectrum disorder (Aylott, 2000). The autism spectrum disorders are a cluster of
different conditions that are sometimes associated with medical conditions (American
Psychiatric Association, 1994). Although there are many different ways in which an
individual might develop autism, the autism spectrum is an umbrella term for many
different symptoms and conditions.

General play behaviours . How do children with autism play? According to
related research, they do not play with toys in a typical fashion, when they choose to play
at all (Tilton & Ottinger, 1964). These children will not remain engaged with any toy or
play object for long periods of time, unless it is one to which they have developed an
attachment (Siegel, 1996). Rarely will they develop an attachment to a particular toy;
rather they prefer to play with other items they have found. Usually these items are just
junk found around the house (e.g., water bottle caps, elastic bands, etc.) that are not often
used as playthings (Siegel, 1996; Winzer, 1987).

It is common for children with autism to choose play objects based on the
sensory stimulation they provide (i.e., interesting to smell, touch, hear, taste or see). For
example, a blanket might be chosen for its softness and its silky edges. The sensorimotor
pleasure the blanket brings will be the reason why the child with autism will form an
"attachment" to it. Specific parts of objects appear to hold the child with autism's


20

attention, as though he or she is not capable of viewing the object as a whole. The object
is not regarded as an item that has a particular function; rather it is a source of sensory
stimulation (Siegel, 1996; Tilton & Ottinger, 1964). For instance, the cap from a water
bottle may be chosen because it can be spun like a top. Sometimes this object will serve
as the child's secure base of attachment. This means that the child prefers to relate to the
object rather than other people (Black, Freeman & Montgomery, 1975; Winzer, 1987). In
fact, when children with autism get upset, they may find this item more comforting than
the presence of their primary caregivers (Siegel, 1996). Children with autism will often
remain in the sensorimotor play stage for longer than other children. Rarely do they use
toys symbolically or combine different toys (Tilton & Ottinger, 1964). This is probably
due to a combination of slower development and an orientation that is more sensory in
nature than that of typically-developing children (Black et al., 1975; Siegel, 1996).

Dolls and stuffed animals are seldom chosen as playthings. It has been suggested
by Siegel (1996) that this is because these toys are supposed to represent real animals and
people and the child with autism may not be capable of taking the "perspective" of his or
her toys. For instance, whereas typically-developing children will pretend their dolls are
hungry or thirsty, it seems not to occur to children with autism to "feed" their dolls
(Siegel, 1996). Being able to take the perspectives of others is what is known in
developmental research as "theory of mind" (for reviews on theory of mind and autism
see Baron-Cohen, 1993). Some researchers believe that children with autism lack a
theory of mind, which is why they experience difficulties with play (Baron-Cohen,
Leslie, & Frith, 1985).


21

In a recent review conducted by Charman (1997), it was determined that some
researchers believed that theory of mind is impaired in children with autism due to
deficits in joint attention and pretend play. Others don't believe that joint attention and
pretend play have a special role in theory of mind; rather the impairments in theory of
mind are symptomatic of a greater cognitive deficit in attention or executive function
(Charman, 1997).

For the child with autism, interaction with a play object does not possess the
qualities of usual play. It lacks the imagination evident in the play of typically-developing
children. For instance, children with autism will line up their dolls but they will not
interact with them or use them to tell a story (Siegel, 1996). The play of a child with
autism has a very repetitive and stereotyped quality, a phenomenon known as
perseveration (Hughes, 1998 ). Perseveration of a sensory nature appears to increase with
retardation in the child with autism. Once engaged in a perseverative behaviour, it is
extremely difficult to interrupt a child with autism (Siegel, 1996). The repetitive quality
of play in children with autism differs from the repetitive play of young people with
typical development in that the repetitions are greater in number and they go on longer.
They have almost a "broken-record" quality to them. For instance, whereas a typically-
developing child might press a button in a book three or four times to hear the
corresponding sound made by the animal pictured there, a child with autism might press
the same button continuously for thirty minutes without paying any attention to anything
else (Siegel, 1996). The repetitions might also include the shaking, spinning and twirling
of toys (Tilton & Ottinger, 1964, Winzer, 1987). Although these findings are interesting


22

and add to a growing body of literature on autism, they appear to be based on case
studies, rather than naturalistic observations (particularly those of Siegel, 1996) . The
present study is an attempt to address the dearth of play research in children with autism
in naturalistic settings.

Cognitive levels of play . What cognitive levels of play do children with
autism typically exhibit? Answers to this question are often contradictory in nature.
Contradiction is particularly evident when reviewing the literature on the cognitive levels
of play engaged in by children with autism. For example, Lewis and Boucher (1995)
found that children with autism did not differ from their peers with learning disabilities
and younger, typically-developing children in terms of their ability to produce ideas for
functional and symbolic play. Similarly, other researchers did not detect a difference
between young people with autism and "mentally handicapped" children in ability to
produce functional play and appropriate object substitutions (Charman & Baron-Cohen,
1997). This research was supported by Rogers (1988), who believed that all children with
developmental delays, including those with autism, demonstrate a fairly typical
progression in sensorimotor play. The difficulties that do exist appear to be in the areas of
motor and gestural imitation. Rogers differed from the other researchers in that she
believed children with autism experience more difficulty than their peers in symbolic
play. She stated that children with autism spend more of their time in immature forms of
play than peers of an equivalent verbal mental age (Rogers, 1988).

Yet, other researchers have concluded that children with autism produce less
functional play and object substitutions when prompted than do children with


23

developmental delay and typical development (Charman, Swettenham, Baron-Cohen,
Cox, Baird, & Drew, 1997).

Conversely, Hughes (1998) stated that children with autism are less likely to
engage in symbolic or exploratory play unless they are cued by an adult. This research
was supported by Riguet, Taylor, Benaroya and Klein (1981), who reported that adult
modeling of symbolic play eUcits higher levels of play in children with autism, although
the modeling usually results in a literal imitation of the adult's behaviour. In addition,
Riguet and her colleagues detected a deficit for imitation in the participants with autism,
as compared with children with Down Syndrome and typical development.

There are two rivaling hypotheses as to the lack of symbolic play in children with
autism. The symbolic deficit hypothesis describes people with autism as incapable of
taking the perspectives of others (see above section on theory of mind). Conversely, the
conative hypothesis suggests a lack of motivation to represent and symbolize as the cause
of the dearth of symbolic play. Proponents of this theory believe that the ability to
symbolize exists in these children because they are able to do so when directly prompted
to (Hughes, 1998).

It should also be noted that the environment may or may not affect the play
behaviour of children with autism (Black et al., 1975). For instance, within a confined
space with no objects present, children with autism are more likely to engage in solitary
repetitive behaviour than in a more open space with play objects available to them.
Aggression and other negative behaviours also occur more frequently. When playing in a
confined space with someone modeling appropriate behaviour and an organized game.


24

children with autism modeled and imitated more frequently than they would if no one
was modeling appropriate behaviour or structuring their time. They also engaged in more
gross motor play when someone was present than when they were alone (Black et al.,
1975).

Social levels of play . When reviewing the literature on social play, it is often
noted that children with autism have difficulties interacting socially with peers (Lord,
1984; Tilton & Ottinger, 1964; Winzer, 1987). Children with autism engage in a minimal
amount of social interactions. It seems as though they prefer to be alone. In fact, they
may become upset when told to interact with others (Winzer, 1987). Siegel (1996)
suggested that the ability to relate well to others is what separates a diagnosis of PDD
from a diagnosis of autism. While children with autism and PDD have trouble relating to
others, children with PDD seem to want social contact. Children with autism appear
uninterested in or unable to relate to others. When children with autism and children with
PDD do interact with others, it is often for self-serving purposes. It is as though it does
not occur to them that others have needs and feelings as well (Siegel, 1996). It is believed
that this is because children with autism have trouble viewing situations from the
perspectives of others (i.e., they lack a "theory of mind"). If they do not understand what
others want, they cannot engage in socially appropriate behaviour with them (Baron-
Cohen et al., 1985).

Lord (1984) investigated the social behaviour of children with autism in play
situations. Although the study was conducted in order to evaluate a specific intervention
procedure, some information was gathered on how children with autism interact with


. <h'


25

typically-developing children during a baseline stage of data collection. When
participants with autism were placed in a different classroom with two unfamiliar
typically-developing same-sex peers (one older and one younger), it was found that the
children with autism interacted with their typically-developing peers at very low rates. In
addition, they played at a great distance (i.e., over 4ft.) from each other. It was very rare
for the participant with autism to initiate interactions with the young person with typical
development. Overall, children with autism either avoided peers altogether, responded
inappropriately to their social invitations, played in an uninterested fashion or initiated
play improperly (Lord, 1984).

Problems with current research . According to Hughes (1999), there are a few
reasons why a dearth of research exists regarding the play behaviours of children with
developmental delays in general. First, much of the research is based on the behaviours of
solitary participants in laboratory settings. This does not make it possible to observe
social behaviours or play in a naturalistic setting. Second, it is difficult to make
generalizations about one diagnostic group (e.g., autism) based on the findings about
another diagnostic group (e.g., Down Syndrome). Third, many children suffer from
more than one developmental delay, making it difficult to determine which disability is
contributing to which play deficit. Finally, like typically-developing children, individual
differences exist among children with special needs. This means that, although a group of
children might all have the same disorder, they will still function at different levels.

In the present study, I attempted to address some of these issues by determining
what types of play behaviour children with autism engage in during free play at school.


;^1 it..


26

Participants were observed in their naturalistic school environments, in the presence of
peers they saw on a daily basis. This was done to insure that an opportunity for social
interaction existed. The study adds to Lord's work because children were examined in a
naturaUstic setting of the classroom, as opposed to a separate room. It was hoped that
confounds, such as adjusting to a new environment and new playmates, would be avoided
by taking this approach. In addition, the typically-developing participants in Lord's study
volunteered to participate (Lord, 1984). This might mean that the typically-developing
participants were more willing to help children with autism than those typically-
developing children who did not participate. A naturalistic classroom setting might
provide the participants with a more realistic situation, where they will encounter
children both willing and unwilling to interact with them. Finally, specific cognitive
levels of play by children with autism were not noted in Lord's study. The current study
included this information. Only children with a diagnosis of autism spectrum disorder
participated. If the participants had diagnoses of additional disorders, parents were asked
to list them.

Why do children with autism and typical development differ in terms of play
behaviour? Several theories have been put forth. One possible explanation is that the
opportunities for play granted to the child with autism by parents and teachers are not the
same as those afforded the typically-developing child (Hughes, 1999). Perhaps caregivers
of children with special needs are more intrusive and/or protective than caregivers of
children without special needs. Children with autism may be perceived as not always as
capable offending for themselves and/or as needing more instruction than their typically-


27


developing peers. Thus, children with autism may be kept closer to home, decreasing the
quality and quantity of their spontaneous play (Hughes, 1999).

It also has been suggested that caregivers may not believe children with
developmental delays are capable of play. These children may experience neglect or may
be permitted only to play with others like themselves. This may occur despite the fact
that previous research supports the view that the optimal play environment for disabled
children is one including playmates who are not disabled, are younger in chronological
age and possess the same developmental age (Bednersh & Peck, 1986).

However, these researchers do not take into account the cognitive difficulties the
child with autism experiences. Children with autism may not be able to communicate
their needs to their peers and may find interactions with other children frustrating. Thus,
they may avoid the situation altogether (Hughes, 1999). In addition, despite the best
intentions and efforts of the child's caregivers to socially integrate him or her with peers,
the peer group may not accept the child. They may perceive the child with autism as odd
or frightening and perhaps will not be motivated to play with him or her. Children with
autism might enjoy play that typically-developing children think is strange or "babyish."

It would be interesting to determine whether parents and teachers differ in their
perceptions of children with autism' s capability of different social and cognitive types of
play. If there is a discrepancy between the parents' and teachers' opinions, the children
may be treated differently at home than at school. Perhaps they are not challenged
enough to play or perhaps they have been given tasks that are too difficult to accomplish,
and they have lost the motivation to try. By noting this discrepancy, it is possible that


28


parents and educators will want to communicate more often on how best to meet the
developmental needs of young people with autism.

Summary

The many different definitions, theoretical approaches and cognitive and social
levels of play make it difficult to identify exactly what play is. For the purposes of the
proposed study, play will be operationalized as an activity which is intrinsically
motivated, freely chosen, pleasurable, nonliteral, and actively engaged in (Rubin et al.,
1983). It is apparent that there is a relationship between specific aspects of the wider
concept of play and typical intellectual, emotional and social development, although the
causal direction of this relationship is difficult to interpret. Is it then possible that children
whose development is not typical miss these important developmental milestones? In
order to answer this question in future research, basic information must be gathered first.
For instance, what types of social and cognitive play does a child with autism usually
engage in when given time to play freely at home or at school? On the following pages, I
will discuss the nature of the proposed study and the research questions it will attempt to
answer.

Rationale

The purpose of the proposed study is to determine what types of play behaviour
the child with autism engages in during free play time at school. This research is
important to conduct because there is a general dearth of research in this area. Although
many researchers have examined the differences in play behaviours between young
people with autism and children belonging to other diagnostic groups (e.g., typical


29

development, Down Syndrome, developmental delay, learning disabilities, Charman &
Baron-Cohen, 1997; Charman et al., 1997; Jarrold, Boucher & Smith, 1996; Lewis &
Boucher, 1995; Libby et al., 1997; Libby et al, 1998), the participants were all examined
on an individual basis in a laboratory setting. Virtually no work has been done on how
these children interact with each other in groups or on how they behave in a naturalistic
setting. This research would be important to carry out in order to compare children with
autism and typical development in terms of types of toys chosen, kinds of play behaviour
engaged in, and whether or not any social interaction takes place between the children
with autism and the young people with typical development.

It also is important to assess participants' play behaviours from the perspective of
those adults who are quite familiar with the participants, such as educators and parents,
thus obtaining some measure of their play in different contexts. Each participant's
teacher, educational assistant and parent(s) was asked to answer a series of questions
requiring them to describe the participant's play behaviours at home and at school. In
addition, the responses from the parents and educators of each child were compared to
determine the consistency of viewpoints on the child's capabilities. Large discrepancies
could indicate an environment that is less than ideal for the child's social, emotional and
intellectual development through play. Thus, perceptions of each participant's play
capabilities and the resources available for him or her will be assessed.

In summary, the research questions this study attempted to answer were: During
free play time at school, at what cognitive and social levels do children with autism play?
What are teachers' and parents' perceptions of participants' play behaviours? Are there


;•(,<;■


30

differences between parent and teacher reports of participants' play abilities? Finally, do
observed play behaviours differ from educator and/or caregiver perspectives? It is hoped
that the answers to these questions will be useful for implementing programs designed to
encourage children with autism to play more frequently and at more sophisticated
cognitive and social levels. It is anticipated that intervention through play will aid
children with autism in many facets of intellectual, social, and emotional development.

METHOD
Participants

Eighteen children with an autism spectrum disorder participated in the
study, although one child was dropped from the study because his chronological and
verbal mental ages were significantly different from the rest of the sample (i.e.,
chronological age of 10 years, 8 months and verbal mental age of 11 years) and his
diagnosis was Asperger's Syndrome rather than autism. Data for the remaining seventeen
participants were analyzed and discussed. Although it would have been useful to include
a control group, it was not possible to do so due to time constraints. In all but three of the
cases, it was the child's mother who provided sociodemographic background information
about the participant. Table 1 summarizes the chronological and verbal mental ages, sex
and diagnosis of each participant. The participants ranged in age from 4.25 to 8.25 years.
The mean age was 6.21 years (SD = 1.22). Seven of the children (41.2%) were enrolled
in senior kindergarten at the time of the study. The remaining children were mainly in
junior kindergarten (23.5%) or Grade 1 (23.5%). Over half (52.9%) of the participants
were raised with one sibling and 29.4% were raised with two siblings. Siblings were


31


Table 1

Sex, Diagnosis. Chronological Age and Verbal Mental Age (as Assessed by PPVT-R) of

Participants

Participant Sex Diagnosis Chronological Verbal mental

Age age


M

Autism

5-4

4-2

F

Autism

7-0

3-11

M

Atypical Autism

7-11

7-5

M

PDD

5-7

2-1

F

No diagnosis

5-0

4-1

M

Autism

4-3

~

M

Autism & PDD

5-7

4-1

M

Autism

5-8

3-7

M

Autism

6-0

3-10

M

Autism & PDD

6-0

4-3

M

Autism & PDD

5-6

~

M

Autism

8-2

—

F

Autism

7-1

4-0

M

PDD

5-6

4-5

F

Autism & PDD

8-3

~

M

PDD

7-8

6-0

M

Autism & PDD

5-1

2-4

Note. Those

participants without

a listed Verbal Mental A

ge were una


the PPVT-R. The PPVT-R is a proxy for verbal mental age, not a mental age
measure.

typically older than the participants. Seven children had a younger sibling either in

addition to, or instead of, an older sibling. One child had a twin and two were only

children. Seven of the parents described their children as possessing a Canadian cultural

or ethnic background. English was the only language spoken in all of the participants'

homes.

Table 1 also indicates the diagnostic categories of the participants. Sixteen of the


r^


32

children had been diagnosed with a disorder on the autism spectrum before participation
in the study. Seven children (41.1% of the sample) had been diagnosed with autism,
five received a combined diagnosis of autism and PDD (29.4% of the sample) . One
child had not received an official diagnosis at the time of the study but was included in
the research because her data was not outlying as compared with the other participants'.

The participants received their diagnoses between the ages of one and five years.
The mean age at which the children were diagnosed with an autism spectrum disorder
was 2.96 years ( SD = 0.89). Six of the participants received an additional diagnosis to
autism. Some examples include ADHD (attention deficit hyperactivity disorder), PDD
and CAPS (central auditory processing deficit). One parent declined to mention what was
the additional diagnosis. It was assumed based on parents' comments that those children
who had a diagnosis before the study commenced received their official diagnoses of an
autism spectrum disorder by appropriate professionals who could provide diagnoses in
Ontario.

Participant verbal mental age (as measured by the PPVT-R) is also listed in
Table 1. Verbal mental age was measured in order to determine at what level the
participants were fiinctioning. This was important to establish so that descriptive
comparisons with typically-developing children of the same approximate verbal mental
age on play behaviours could be made. The participants were asked to examine pages in a
book. Each page contained four pictures. The children were told to identify a specific
picture on each page.

Four children did not complete the PPVT-R, two because they got upset when


33


they were asked to participate and two because they were not yet able to point
independently, which the test required. The mean verbal mental age for the participants
was 4.39 years (SD = 1.60). A paired-samples t-test revealed that participant verbal
mental age and chronological age differed significantly (t (i2)= 4.86, p < .0001). All of
the participants were functioning at a level of at least 6 months below that of their
chronological age. This may indicate a delay in receptive vocabulary ability and perhaps
a delay in academic functioning.

The parents were asked to indicate what was the nature of their children's
programs at school. All of the children had an educational assistant (EA) present with
them in the classroom. Seven of the children attended half-day sessions at school and
seven attended fiill-day sessions at school. The remaining three children attended half-
day sessions at school and half-day sessions in daycare or a special program.

Fourteen (82.4%) of the participants lived with both parents at the time the
study took place. The rest of the children lived either with one parent or in a multi-
generational setting. In more than half (64.7%) of the cases, the children's mothers
stayed at home with them when they were not at school. Either parent cared for four of
the participants after school, depending on who was not working. One child was looked
after by his/her father, and one by his/her grandparents. In most (10.6%) of the cases,
the mother met regularly with the participant's teacher or EA. In the remaining cases,
both parents met with their child's educators.

As reported by the participants' mothers and the one father who filled out the
background information, the educational backgrounds of each child's mother and father


34

(where applicable) were also indicated. Eight (47%) of the mothers and seven (41.2%) of
the fathers had completed either a university degree or a college diploma. Nine (52.9%)
of the fathers and seven (41.2%)) of the mothers had completed high school. Only one
mother and one father listed junior high school as the highest level of education they had
completed. One mother had completed her M. A.

Five participants were present for the pilot testing of the scale to be used in the
naturalistic observation and 12 participated in the actual study. The children resided in
the Niagara region of Ontario, Canada. I sought permission from the Niagara Catholic
District School Board and the District School Board of Niagara to contact the
participants' schools. Eleven schools from the Niagara Catholic District School Board
and four from the District School Board of Niagara participated in the study. Next,
permission was granted by each school's principal for me to try and recruit participants,
their parents and their teachers. Participants' names were either released by the school
board or the principal contacted the parent first. Parental and child consent were required
before each child could participate in the study. Thirty-four families were invited to
participate in the study. Eighteen agreed to participate, and sixteen declined participation.
Procedure

Pilot testing of the Play Observation Scale (POS) occurred before the study
commenced. Five participants were observed as they played during five separate free
play periods at school. Categories were either added to, or removed from, the POS based
on what was observed during pilot testing.

The study took place in the schools of the children with autism, by permission


35

of the local school boards, the schools' principals and the children's parents and
educators. Teachers introduced the children to myself and the reliability observer (where
applicable) upon our first visit, and were asked to remind the children of our presence and
purpose for each subsequent visit to the schools. Participants were observed as they
participated in free play period in their mainstreamed classrooms. Depending on the
child's age, this was either during center time (kindergarten) or recess (Grades 1-2). Each
participant was observed a minimum of five times, to counteract for effects such as
participant/teacher mood, environmental changes, etc. Each session lasted for
approximately 10 minutes of the class' free play time (i.e., five minutes' worth of coding
and five minutes' worth of detailed note-taking on what the child was doing), with the
exception of the first session, when we were present for an additional 10 minutes in order
to get each child accustomed to our presence. Every 10 seconds, we recorded the child's
cognitive and social level of play as well as any transitional, unoccupied or onlooker
behaviour observed during the play period. Anxious behaviours, aggression, and
perseveration if displayed by the participant also were recorded. This coding procedure
is based on previous studies in which the POS was used (Coplan, Rubin, Fox & Calkins,
1994; Hymel, Rubin, Rowden & LeMare, 1990; LeMare & Rubin, 1987; Rubin et al.,
1976; Rubin etal, 1978).

Inter-rater reUability was calculated on the POS to insure that data was being
coded based on objective operationalizations rather than subjective criteria. The second
observer was present for 20% of the observations. Inter-rater conflicts were resolved by
discussion immediately afl;er each coding session.


36


Reliability analyses revealed a Cohen's Kappa of .86 (p < .001). Inter-rater
agreement was 86.6%. Cohen's Kappa for nominal scale agreement was selected for a
couple of reasons. First, unlike the percentage agreement procedure, it displays all
agreements and disagreements between observers. This makes it easier to train a second
observer, as specific disagreements can be noted and discussed. In addition, it can be
determined whether an error occurred as a result of omission or commission. Second,
Cohen's Kappa was used because it corrects for chance agreement by removing it from
the analysis. Other reliability procedures (e.g., percentage agreement, Scott's coefficient)
do not consider chance agreement (HoUenbeck, 1978).

Each participant's teacher and/or educational assistant and parents were also
requested to fill in a brief series of questionnaires asking them to describe the
participant's play behaviours. Educators were interviewed during spare time at school,
and parents were interviewed in the child's home at a convenient time for them. If
possible, both parents were interviewed. Otherwise, the parent who was available was
invited to participate in the study. Sociodemographic information was only obtained from
the parents. In addition, the parents and educators were verbally interviewed, using a
series of numerically-coded and open-ended questions about the child's play capabilities
and the resources available for him or her. During the visit to the participant's home, I
also determined the child's verbal mental age, using the PPVT-R. The visit to the
participant's home usually took approximately 11/2 hours and the educator interview
took approximately 40 minutes (20 minutes for the teacher and 20 minutes for the EA).


37

Measures

Background information . The participants' parents or guardians were asked to
provide demographic variables such as participant age, diagnostic category (i.e., type of
Pervasive Developmental Disorder), year of diagnosis, ethnic background and schooling
information. They were asked which primary caregivers their children lived with, their
relationship to the child, how care is arranged for the child, which parent is responsible
for the majority of the childcare, whether the participant has sibHngs, and the highest
level of education completed by the caregiver participating in the study and the
caregiver's spouse.

Verbal mental age . The Peabody Picture Vocabulary Test-Revised (Dunn &
Dunn, 1981) was used to assess verbal mental age. The PPVT-R measures receptive
vocabulary in children, as opposed to general intelligence, and vocabulary is strongly
associated with academic success. The test does not require the research participant to be
able to read or write, which makes it ideal to use with children who experience language
difficulties, such as those with autism. The PPVT-R was standardized on a representative
sample of American children, youth, and adults (Dunn & Dunn, 1981).

During the administration of the PPVT-R, the participant is required to look at a
series of pages containing groups of pictures. The experimenter asks the child to point at
individual pictures on each page. Correct and incorrect responses are marked on a Test
Record sheet. The raw score is then calculated by subtracting the number of errors from
the number of correct items within the range of ability for children of the same
chronological age. Next, the raw score is converted to a derived score representing age


38


norms (i.e., standard score equivalents, percentile ranks, stanines and developmental-type
age norms such as age equivalents, Dunn & Dunn, 1981).

This measure has been used previously with children with autism as an estimate
of intellectual functioning (McCabe et al., 1996; Morgan, Cutrer, Coplin & Rodrigue,
1989). Morgan and colleagues (1989) used the PPVT-R with children with autism of a
similar chronological age to the current sample, and they found it easier to administer
than more comprehensive tests, such as the Stanford-Binet Intelligence Scale (Terman &
Merrill, 1960) or the Weschler Preschool and Primary Scale of Intelligence (Weschler,
1967). In addition, the PPVT-R correlates moderately highly with both of these scales (r
= .62 and r = .58, respectively, Dunn & Dunn, 1981).

The reliability of the PPVT-R appears to be satisfactory. The internal consistency
of the scale (child and youth version) ranges from .67 to .88, with a median of .80. The
test-retest reliability ranges from .73 to .91, with a median of .82. The non-revised
version of the PPVT correlates most highly with vocabulary measures (e.g., Full-Range
Picture Vocabulary Test, r = .86), correlates moderately well with other tests measuring
verbal intelligence (e.g., WAIS Verbal Scale, r = .71), and correlates to a reasonable
degree with measures of school achievement (e.g., Peabody Individual Achievement
Test, r = .68, Dunn & Dunn, 1981).

Play information provided by parents and teachers . The adults responded to a
brief written questionnaire asking them to describe the play behaviours of their
children/students. The Preschool Play Behaviour Scale (PPBS: Coplan & Rubin, 1998) is
a teacher-rating scale created to examine the social and cognitive play behaviours of


ii:*f^


39


young children. For the purposes of this study, the scale was utilized for parental, in
addition to teacher, use. The adults were asked to indicate how often the child engages in
different types of play behaviours. The PPBS is a five-point Likert scale (ranging from
never to very often).

The PPBS consists of five subscales (Coplan & Rubin, 1998). Reticent behaviour
is a combination of onlooker and unoccupied behaviours (i.e., staring into space,
watching others play, but not joining them, etc.). Solitary passive behaviour consists of
playing alone while trying to explore objects, examine objects, construct objects, etc.
Solitary active behaviour can be defined as playing alone while engaging in activities for
the physical sensations they yield (i.e., motor activities, pretend play, etc.). Social play
can be operationalized as peer conversation, group and sociodramatic play. As its name
implies, rough-play consists of rough-and-tumble play and mock fighting.

I added a sixth category to the PPBS, labeled as the "autism subscale." The items
on this subscale described behaviours believed to be typical of children with autism when
playing. The adult participants were asked how frequently the children played in a
repetitive fashion, preferred junk items to appropriate toys, preferred to play with the
same toy all the time (even when other toys were offered), lost interest in new toys
quickly, chose toys based on the sensory stimulation they might provide and/or played
more appropriately when prompted to do so.

Construct validity for the PPBS was established by Coplan and Rubin (1998) by
examining the relation between teacher-rated (i.e., on the PPBS) and observed child
behaviours. In general, moderate to high correlations were found between observed and


40

teacher-rated behaviours (Coplan & Rubin, 1998). Discriminant validity was assessed by
computing correlations between teacher ratings of reticent, solitary passive and solitary
active behaviours and observed non-corresponding nonsocial behaviours. Overall, these
behaviours indicated good discriminant validity (Coplan & Rubin, 1998). The internal
consistency of the subscales was satisfactory with alpha values for the subscales at .96 for
social play, .85 for reticent, .79 for solitary passive, .92 for rough play, and .76 for
solitary active behaviours (Coplan & Rubin, 1998).

The Cronbach alphas for each subscale in the current study (for the combined
caregiver and educator ratings) are listed as follows: Solitary active play = .85, solitary
passive play = .70, reticent behaviours = .59, rough play = .70, social play = .92. A
generally acceptable level of significance for reliability is .70 (Nunnally, 1978). A
possible explanation for the lower alphas for some of these subscales is that the PPBS
was used to discuss play behaviours in typically-developing children at the preschool
level (Coplan & Rubin, 1998), while the participants in the current study were children
with autism. It would appear that there was more variability within the current sample
than there was in Coplan and Rubin's (1998) sample. Furthermore, some of the subscales
were based on only a few items (e.g., rough play was made up of two items). It is
possible to have a lower alpha when only a small number of items make up a subscale
(Nunnally, 1978).

In addition, an autism subscale was added to code specific behaviours associated
with autism, such as perseveration and play with junk items. The Cronbach alpha for this
subscale was .33. The alpha was probably low because the items on the scale were not


41

highly related. For example, although the children played more appropriately when
prompted quite frequently, they almost never played with junk items. When item # 24 on
the scale, "Plays more appropriately when prompted to do so," was removed, the
Cronbach alpha for the autism subscale became .62. Thus, this item was left out in all
analyses.

The participants' parents and educators also were invited to complete a brief
interview with the researchers. They were asked if the participant engages in each social
and cognitive level of play under investigation (i.e., sensorimotor, functional, symbolic,
sociodramatic, constructive, games- with-rules, solitary, parallel, and group). They
responded to these questions using a four-point Likert scale ranging from no to rarely to
sometimes to yes. I also inquired about the participants' non-play behaviours while in a
free play situation (i.e., whether or not he/she remains occupied during play time, if
he/she watches what others are doing without getting involved, is aggressive, etc.). The
parents and educators were given the opportunity to give the researcher any other
information about the child's behaviour that might have been overlooked.

The next part of the interview concerned the type of toys preferred by the
participant. The responses to some of the questions were open-ended. Those requiring a
yes or no answer were coded using a three-point Likert scale. The possible responses
ranged from no to sometimes to yes. I asked the parents and educators to describe the
participant's favourite toy and the type of sensory stimulation they believed it may
provide the child. They were also asked to explain how the child plays with the toy (i.e.,
whether or not his or her actions are repetitive, how long he or she plays with it, etc.) and


42

whether or not the participant gets upset when the toy is removed from him or her. In
order to determine how attached the child is to the toy, the participant's educators and
parents were asked if that particular toy is sought out by that child to comfort him or her
when he or she is upset.

The subsequent section of the interview addressed the participant's play
behaviours in social settings. The parents and educators were asked if the child enjoys
playing with his or her peers and if he or she has a favourite playmate. I also asked how
the children usually play together (e.g., list the kinds of activities they enjoy). Finally, the
participants' parents and educators were asked if the child ever grows upset when asked
to play with others and what kinds of things must be done to calm the child down.

The last section of the interview dealt with the role of the parent and educator in
the participant's play experiences. The adults were asked if they believed that the
children with autism had the same opportunities to play as their typically-developing
peers. They were also asked to describe their roles as influential adults in the child's daily
play times.

Play scale . I used the Play Observation Scale (POS, Rubin & Coplan, 1998) to
record both the cognitive level of play observed and the type of social participation
engaged in by each child. Since this scale was originally intended for typically-
developing children at the preschool age in a laboratory setting, some modifications were
required before it could be used with children with autism in higher grades in a
naturalistic setting. In order to ensure that no behaviours had been forgotten or needlessly
included in the modification of this checklist, as well as to establish the appropriateness


43

of maintaining more complex play behaviours on the checklist (e.g., group-dramatic
play), the modified play scale was pilot-tested before it was used in the study. I observed
five participants in order to determine what types of behaviours should be included in the
checklist. Results of the pilot-testing will be outlined below.

Participants' play behaviours were observed naturalistically every 10 seconds
for five minutes, yielding a total of 30 observations for each play period. The observers
simply checked off whatever behaviours applied fi"om a checklist. In order to avoid
coding too many social and cognitive behaviours within one ten-second interval, the
behaviour that occurred most fi-equently was coded. For example, if a participant engaged
in parallel-constructive play for six seconds and solitary-fianctional play for four seconds,
"parallel-constructive," would be coded.

Table 2 summarizes the definitions of the categories used on the modified POS.
During pilot testing, all of the original categories listed by Rubin and Coplan (1998) were
included in the scale. Rubin and Coplan referred to Smilansky's (1968)
operationalizations of cognitive levels of play and Parten's (1932) social play schemas
when they constructed the POS. They nested the cognitive play items within the social
categories when they created the scale (Rubin & Coplan, 1998).

The checklist was structured in such a way as to allow the researcher to check off
both the cognitive and social levels of play engaged in by the participant. For example,
the solitary play categories consisted of solitary-functional, solitary-constructive, etc.
Rubin and his colleagues (1976) argued that group-fiinctional play and solitary or
parallel-games- with-rules, could not occur due to the nature of the definitions of the


I''


44


Table 2

Variables Measured by the POS and their Operationalizations


Variable


Operationalization


Example


Functional play


Play with an object for the
fiinction it denotes;
repetition of movements for
the pleasure they bring.


Running a toy train along a
track, running around.


Constructive play


Dramatic play


Games-with-rules


Exploratory play


Solitary play


Parallel play


Group play


Building or creating
something to satisfy
constructive goals.

Make-believe and pretend
behaviours, role playing.

The creation of games with
rules and roles to be played
agreed upon the group.

Examining a play object in
order to determine its
properties or how it works.

Playing apart from other
children at a distance of at
least three feet or with one's
back to other children.

Playing within three feet of
others, using similar
materials, but not playing
with them.

Playing with others with a
common purpose to the
activity; interacting
socially.


Building a sand castle,
solving a puzzle.


Creating a puppet play,
dressing up as a firefighter.

Board games, tag, sports.


Putting Lego in mouth or
running fingers along block
instead of playing with it.

Solving a puzzle alone and
away from others while
other children are playing a
board game together.

Solving a puzzle alone and
near others, while other
children are solving a
puzzle together.

Solving a puzzle with other
children.


45


Table 2 Continued

Variables Measured by the POS and their Operationalizations


Variable


Operationalization


Example


Transitional behaviour


Unoccupied behaviour
Onlooker behaviour


Peer conversation


Anxious behaviours


Hovering


Rough-and-tumble play


What the child is doing
after completing one play
activity and before starting
another.

The child is not focused on
one activity in particular.

Watching the play of others,
and perhaps engaging in
conversation with them, but
not trying to join them.

The verbal interaction
between children in a play
situation.


Behaviours a child might
engage in when upset or
nervous.


Not playing, but being
within less than three feet of
other children who are
playing, for at least three
seconds, then moving to
another group. Could be
watching several activities
at once.

Play fighting, with no
intention to harm others.


Cleaning up a finished
activity or gathering
materials for a new activity.


Wandering aimlessly or
staring off into space.

Watching peers play
baseball, observing peers
playing a computer game
and making comments.

One child shows a new toy
to another. The other child
asks questions about the
toy.

Automanipulatives (thumb-
sucking, pulling on one's
own clothes or hair), crying,
self-stimulation.

Watching one group play
and then moving to another
group after less than three
seconds. Continues to do
this for the entire ten-
second interval.


Wrestling, chasing and
hitting while laughing.


46


Table 2 Continued

Variables Measured by the POS and their Operationalizations


Variable


Operationalization


Example


Aggression


Perseveration


Junk object play
Leaving the classroom


Adult interactions


Socially inappropriate
behaviour


Self- or toy-directed talk


Participant-initiated play


Intent to harm another
child, verbally, physically
or emotionally.

Constant, lengthy and
numerous repetitions of a
particular behaviour.


Play with items that are not
considered toys.

The participant walks away
from the observation area
and it may not be possible
to follow him/her.

Interacting with the teacher,
educational assistant or the
researcher.

Behaviour that is prohibited
by the rules of the
classroom.

Vocalizations made by the
child that are directed to a
toy or to him or herself

The child invites a peer to
play with him/her.


Name-calling, punching,
kicking, fighting over a toy.


Repeatedly pressing a
button on a computer for
the sound it makes.
Repetition might continue
for a long time.

Playing with string, elastic
bands, lids for water bottles.

Going to the washroom,
lockers or the principal's
office.


Gives EA a hug, asks the
researcher what she is
doing.

Standing on furniture,
playing with teaching
resources instead of toys.

Muttering to self, telling a
toy dinosaur to run.


Brings a puzzle to another
child and asks that child to
join him/her.


47


Table 2 Continued


Variables Measured bv the POS and their Operationalizations


Variable Operationalization Example


Peer-initiated play


Helping behaviour


Peer redirects/motivates


A peer invites the
participant to play with
him/her.


The participant provides
assistance to a peer.

A peer attempts to stop the
participant from behaving
inappropriately or to
motivate the participant to
play appropriately.


Another child asks the
participant to play and the
participant responds to this
request.

Taking a hurt friend to the
nurse's office.

A peer stops the participant
from wandering off.


cognitive categories (Rubin, Maioni, & Hornung, 1976).

During the pilot-testing of the POS, the category "perseveration," was added to
the scale, due to the frequency with which individuals with autism are believed to engage
in this behaviour when they play (Siegel, 1996). I also added the play level, "junk
object play," to the scale, due to the notion that children with autism often prefer to play
with junk as opposed to age-appropriate toys (Siegel, 1996).

Pilot testing of the POS resulted in some changes in the scale for the purpose of
observing children with autism. The category "junk object play," was removed from the
scale after pilot testing because none of the participants had motive or opportunity to play
with junk items during free play time. Hovering was also deleted from the POS when
pilot testing did not reveal one instance of this behaviour by the children with autism.


48

Moreover, I found it difficult to differentiate between the categories, "hovering," and
"unoccupied."

I observed some behaviours that were not indicated on the checkhst. Leaving the
classroom, adult interactions, socially inappropriate behaviours, laughing, helping
behaviours, peer attempts to redirect or motivate the participant, peer initiated play,
participant initiated play and talking to self or toys were added to the collection of actions
listed in order to accurately account for how free play time was spent by the participants.

RESULTS

In this section, I will first describe the reported play behaviours that the children's
parents engaged in with them. Next, my observations will be summarized according to
how oflen each social and cognitive type of play behaviour occurred. Following this,
parent and educator perspectives on participant play behaviours, as measured on the
Preschool Play Behaviour Scale and the interview, will be discussed and then compared.
Finally, my observations will be weighed against the parent and educator perspectives on
what the participants do during their free play time.
Play Activities Engaged in by Caregivers with Participants

The participants' caregivers were asked what play activities they usually engaged
in with their children. The most commonly listed were reading books (41.18%), sports
(e.g., baseball, hockey, 41.18%), crafts and play dough (35.29%), working one-to-one
(e.g., on schoolwork and concepts, 35.29%), computer games (35.29%) and going for
hikes and walks (29.41%). Respondents were permitted to list more than one activity, and
many listed several.


49


Research Question One: The Cognitive and Social Levels of Play Engaged in

by Children with Autism

The first research question explored in this study was, "During fi-ee play time at
school, at what cognitive and social levels do children with autism play?" To address this
question, I conducted a naturalistic observation of the participants at their schools during
their fi-ee play time (e.g., recess, center time, etc.). I used the Play Observation Scale
(POS, Rubin & Coplan, 1998), a checklist for each social and cognitive level of play
observed during a free play session.
Observed Cognitive and Social Play Behaviours

The social (i.e., solitary, parallel or group) and cognitive (i.e., fijnctional,
exploratory, dramatic, games, or constructive) categories of play each participant was
engaged in was recorded every ten seconds. Any additional behaviours (e.g., crying,
socially inappropriate behaviour, perseveration, snack time, laughing, peer redirects or
motivates, peer-initiated play, participant-initiated play, self or toy talk and left the room)
were also noted where applicable. Table 3 summarizes the observed cognitive, social and
other play behaviours of the participants, ranked according to the percent of total
behaviours and the percent of total time (i.e., 150 ten-second intervals) they represent.
Table 3 contains the social and cognitive behaviours as they were combined during
observational coding (e.g., solitary-fianctional, group-dramatic, etc.). Each of the five
sessions consisted of 30 ten-second intervals.

Parallel-fianctional play occurred most frequently, followed by adult interactions.
Solitary-fiinctional play occurred the next most frequently. Transitional behaviour was


50


Table 3

Percentages of Total Observed Play Behaviours Across all Children in Descending Rank

Behaviour Percent of Total Ten-
^ Second Intervals

Parallel-functional 13.21

Adult interactions 11.53

Solitary-functional 9.92

Transitional 9.33

Unoccupied 8.35

Peer conversation 7.06

Parallel-constructive 7.06

Onlooker 6.63

Group-games 5.37

Parallel-exploratory 4.67

Solitary-exploratory 4.59

Solitary-constructive 3.84

Solitary-games 2.08

Group-functional 1 . 73

Parallel-games 1.49

Group-constructive 1.21

Left room 0.83

Rough-and-tumble 0.67

Parallel-dramatic 0.59

Group-dramatic 0.59

Group-exploratory 0.47

Anxious behaviours 0.47

Solitary-dramatic 0.39

Aggressive behaviours 0.08

Crying 0.08


Note. For percent of total observed behaviours, percentages do not add to 100% due to
rounding error. These categories are exclusive. Due to the nature of the sample's
interactions, the categories, "adult interactions," "crying" and "left room" were added to
the scale.

ranked immediately following solitary-functional play and before unoccupied and

onlooker behaviour. Adult interactions occurred more often than did peer conversations.


51

The least observed cognitive and social behaviours were group-exploratory and
solitary-dramatic play. Aggression and crying were the least observed behaviours in the
participants during free play time. Peer- or participant-initiated activities were removed
from the final analysis because it was sometimes difficult to determine who had initiated
what after having stopped watching for a few seconds to code. Laughing, snack, socially
inappropriate behaviour and self- or toy-directed talk were also removed because they
occurred so rarely. Finally, perseveration also was excluded from the analysis because it
was difficult to operationalize and it occurred so rarely.

Table 4 contains the same observational data rearranged according to the separate
social and cognitive categories. The most frequently observed social play behaviour was
parallel play, followed by solitary play. The most frequently observed cognitive play
behaviour was fiinctional play, followed by constructive and exploratory play. The rarest
social behaviour was group play and the rarest cognitive behaviour was dramatic play.
The participants engaged in adult interactions more frequently than they did peer
conversations. Transitional behaviour occurred more frequently than unoccupied or
onlooker behaviours. Aggressive behaviour and crying occurred the least often, followed
by the anxious and rough-and-tumble categories.

Research Questions Two and Three: Caregiver and Educator Perspectives on

Participant Play Behaviours

The second research question I attempted to answer was, "What are educators'
and parents' perceptions of participants' play behaviours?" To address this inquiry,
participants' caregivers and educators filled out a questionnaire and responded to an


52


interview on the play behaviours of the participants. Differences in parent and teacher

reports of participants' play behaviours also were compared, in response to the third

research question.

Table 4

Percentages of Total Observed Play Behaviours Represented by Each Social and
Cognitive Categorv in Descending Rank

Behaviour Percent of Total Observed Behaviours

Parallel 18.97

Functional 11.42

Solitary 11.06

Adult Interactions 8.23

Constructive 7.95

Exploratory 6.88

Transitional 6.66

Group 6.47

Unoccupied 5.96

Peer Conversation 5 . 04

Onlooker 4.73

Games 4.09

Dramatic 0.98

Rough and Tumble 0.48

Anxious 0.34

Aggressive 0.08

Crying 0.08

* Note . These categories are exclusive. Due to the nature of the sample's interactions, the
categories, "adult interactions," "crying" and "left room" were added to the scale.

Comparing Parent and Educator Perspectives on Participant Play Behaviours

PPBS. The participants' caregivers and educators were asked to fill out a pencil

and paper questionnaire measuring their observations of the participants' play

behaviours. Responses were gauged using a five point Likert scale ranging from never (1)


53

to very often (5).

In order to compare caregiver and educator perspectives on participant play
behaviours, composite scores were created by summing the responses to the questions on
the PPBS corresponding to the six categories of play behaviours measured by the scale
(i.e., solitary active, solitary passive, reticent, rough and tumble, social play and autism-
like behaviours without question # 24).

Since six t-tests were conducted for each adult respondent group (i.e., teachers,
EAs and mothers), a modified Bonferroni test was used to determine the level of
significance to adopt for the t-tests. This procedure was selected because it includes the
degrees of freedom associated with the orthogonal comparisons used in the analysis. By
using this procedure, corrections to the significance level are only introduced when the
number of comparisons exceeds the degrees of freedom (Keppel, 1991).

The main concern about the use of this adjustment for significance is that it may
be too conservative and thus runs the risk of not detecting a true difference. Upon further
examination of the statistical output, however, not many of the t-tests were significant,
even at the .05 level. In fact, the closest p value was .029. Even if significance was set at
.01, the differences between mothers and educators for comparison on "Does your
student grow upset easily when asked to play with others?" would not have been
significant. Hence, I maintain that the modified Bonferroni procedure was appropriate to
use in this situation.

Using the modified Bonferroni procedure, the level of significance was set at p =
.008. All analyses conducted in this study used two-tailed probability values.


54


Because educational assistants (EAs) and teachers based their reports on
observations from the same school environment, the first comparisons were made
between teachers and EAs. The answers provided by the teachers and educational
assistants were averaged together into one set of composite scores after paired-samples t-
tests did not reveal any significant differences between teacher and educational assistant
perspectives, largest t(i2) = 1.80, p = .098, for the autism subscale comparison.

Next, mother and educator (i.e., averaged scores of teachers and EAs)
perspectives of the participants' play behaviours were compared using paired-samples t-
tests. Table 5 summarizes the results of these t-tests. No significant differences emerged
between these two groups, largest t(i3) = 2.37, p = .033, for comparison on the social play
subscale. The responses of the second caregivers (i.e., fathers, grandparents, etc.) were
not included in the analyses because so few of them participated in the study.
Table 5

Summary of Paired-samples T-tests, Means and Standard Deviations for Mother and
Educator Perspectives of Participant Play Behaviours (PPBS)


Mothers


Educators


Variable


T-test


df


Social play

2.38

14

Solitary passive

0.63

13

Reticent

0.55

14

Rough play

1.98

14

Solitary active

2.24

14

Autism

0.72

12


M

2.74
3.57
2.87
2.63
3.53
3.00


SD

0.79
0.78
0.68
1.23
1.03
0.70


M


SD


2.27

0.68

3.45

0.49

2.76

0.43

1.93

0.65

2.73

1.13

2.84

0.56


Note. 1 = never, 2 = hardly ever, 3 = sometimes, 4 = often, 5 = very often.


55


Since there were no significant differences between the mother and educator
perspectives, the mean responses for the two groups were combined and presented in
Table 6.

According to the caregivers and educators, the most frequently occurring
participant play behaviours were the solitary passive behaviours. The cognitive types of
play included in this subscale were exploratory, constructive and fimctional. The least
frequently occurring behaviour was rough play. Social play was also rare.
Table 6

Summary of Paired- samples T-tests. Means and Standard Deviations for Averaged
Mother and Educator Perspectives of Participant Play Behaviours TPPBS)


Variable

T-test

df

M

SD

Social play

2.38

14

2.43

0.62

Solitary passive

0.63

13

3.49

0.50

Reticent

0.55

14

2.79

0.36

Rough play

1.98

14

2.16

0.61

Solitary active

2.24

14

3.00

0.88

Autism

0.72

12

2.89

0.49

Note. 1 = never.

2 = hardl>

' ever, 3

= sometimes, 4 =

= often, 5 = very ^


In order to determine if the caregivers and educators perceived any of the play
behaviours as occurring significantly more frequently than the others, paired-samples t-
tests were conducted. Using the modified Bonferroni procedure, the level of significance
was set atp= .017.

Results of the t-tests are summarized in Table 7. Solitary passive play was
perceived by the participants' caregivers and educators as occurring significantly more


56


frequently than all of the other behaviours (social play, t (i3)= -6.43, g < .0001, reticent
behaviours, t (i3)= 4.42, ^ = .001, rough-and-tumble play, t(i3)= 6.09, p < .0001, solitary
active play, t(i3)= 2.85, p = .014 and autism-like behaviours, t(i2)= 3.53, p = .004).
Reticent behaviours (t(i4)= 3.27, p = .006), solitary active play (t(i4)= -3.39, p = .004) and
autism-like behaviours (t(i2)= -4.35, p = .001) occurred significantly more frequently than
rough-and-tumble play. Rough-and-tumble play was perceived by the caregivers and
educators as the least frequently occurring play behaviour.
Table 7
Summary of Paired-samples T-tests for Comparisons Between PPBS Subcategories


Comparison


T-test


Df


Social vs. Solitary passive

Social vs. Reticent

Social vs. Rough-and-tumble

Social vs. Solitary active

Social vs. Autism

Solitary passive vs. Reticent

Solitary passive vs. Rough-and-tumble

Solitary passive vs. Solitary active

Solitary passive vs. Autism

Reticent vs. Rough-and-tumble

Reticent vs. Solitary active

Reticent vs. Autism

Rough-and-tumble vs. Autism

Rough-and-tumble vs. Solitary active

Solitary active vs. Autism


-6.43*
-1.67
1.38
-2.66
-1.78
4.42*
6.09*
2.85*
3.53*
3.27*
-0.80
-1.20
-4.35*
-3.39*
0.05


13
14
14
14
12
13
13
13
12
14
14
12
12
14
12


Note. *p<. 017

Interview . The caregivers and educators were also verbally interviewed to gather
more information about their perspectives on the play behaviours of the participants. The


57

interview differed from the PPBS in that the respondents had the opportunity to provide
more open-ended responses. Sixteen mothers, seven other caregivers and seventeen
EAs completed the interview. Fifteen teachers were interviewed and two collaborated
with the EAs because they were too busy to complete a separate interview themselves.
The opinions of the additional caregivers are excluded here because so few were able to
participate in the study.

The respondents were first asked to specify whether or not the participants
engaged in a series of different social or cognitive play behaviours. Their responses were
based on a four-point Likert scale ranging from no to rarely to sometimes to yes. Answers
stating that the participant required prompting to engage in a certain behaviour were
excluded from the analysis because I was interested in looking only at the behaviours the
participants were capable of on their own.

The questions were arranged according to the variable they measured (e.g., "plays
with an object without taking its functional features into account = exploratory play).
Teacher and educational assistant responses to these questions were compared using
paired-samples t-tests. Using the modified Bonferroni procedure, the level of
significance was set at p = .005. No significant differences were found, largest t (9) =
-1 .77, p = . 1 1, for comparison on the question, "Playing with others with a common
purpose to the activity." The teachers' and EAs' responses were then averaged and
compared to the mothers', using paired-samples t-tests. Once again, no significant
differences were found, largest t(io) = 2.19, p = .05, for the comparison on how often
participants engage in role-playing. The results of the t-tests comparing mother and


58

educator perspectives on the first set of questions for tlie interview are displayed in Table
8.

It was interesting to note that the mothers were unanimous in saying that their
children engaged in both solitary and parallel play. In addition, both the educators and the
mothers were more likely to state that functional play is engaged in by the participants
than any other cognitive form of play. Although the mothers stated that the children
engaged in dramatic play more than they did exploratory play, the educators observed the
reverse. Finally, the mothers and educators reported onlooker behaviour as occurring
more frequently than unoccupied behaviour.
Table 8

Summary of Paired-samples T-tests. Means and Standard Deviations for Mother and
Educator Perspectives of Participant Play Behaviours (Interview Question 1)

Mothers Educators

Variable T-test df M SD M SD

2.83 1.40


Exploratory

-0.74

11

Functional

1.05

12

Dramatic

2.19

10

Constructive

2.06

10

Games

2.32

3

Solitary

1.99

13

Parallel

1.76

12

Group

1.28

8

Unoccupied

0.43

11

Onlooker

1.63

13

Note. 1 = no, 2 =

rarely, 3 =

som


3.92

0.28

3.64

0.92

3.73

0.91

2.50

1.29

4.00

0.00

4.00

0.00

3.44

1.01

2.54

1.16

3.14

1.29


3.17

0.62

3.73

0.56

2.73

1.10

3.41

0.94

1.75

0.87

3.75

0.47

3.85

0.32

2.72

1.23

2.33

1.12

2.68

1.19


The interview questions requiring yes or no answers (e.g., "Does your child have


m.


59

a particular toy that he/she plays with more often than all the others?" "Does your child
grow upset if this toy should go missing from him/her?" etc.) were coded on a three-point
Likert scale ranging from no (1) to sometimes (2) to yes (3).

Paired-samples t-tests revealed no significant differences between the teachers
and the EAs in response to these questions, largest t(i4) = -2.43, p = .029, for comparison
on "Does your student grow upset easily when asked to play with others?" Once again,
the modified Bonferroni procedure was used, with the level of significance set at p =
.008. The teachers' and educational assistants' responses were averaged and then
compared to the mothers' responses. The results of the t-tests are summarized in Table 9.
No significant differences existed between the educators and the mothers on these
questions, largest t(6)= 1.44, p = .20, for comparison on "Does your child grow upset if
[his/her favourite toy] should go missing or be taken from him?"

The interview questions requiring open-ended responses were examined next.
Respondents were asked if there were any additional behaviours the participants
engaged in that were not mentioned already. Many different play behaviours were
mentioned, but no visible trends were observed. Some examples of responses include
repetition (of movement and toy choices), mimicking other children or dialogue from
television shows and the need to fit in with others.

As indicated in Table 9, the majority of the respondents said that the participants
had a particular toy that they played with more often than all the others. Responses for
both mothers and educators ranged between "sometimes," and "yes." If the respondents
answered "yes" to the question, "Does your child/student have a particular toy that he/she


60


plays with more often than all the others," they went on to answer a series of questions
about this toy. The following percentages are based on those who answered "yes." When
asked to describe the participants' favourite toys, 36.4% of the mothers and 29.1% of the
educators mentioned train sets, with particular reference to "Thomas the Tank Engine."
Constructive toys, such as Lego and blocks were also listed by 9. 1 % of the mothers and
20.7% of the educators.
Table 9

Summary of Paired-samples T-tests. Means and Standard Deviations for Mother and
Educator Perspectives of Participant Play Behaviours (Interview Questions 3. 5. 7 & 8)

Mothers Educators

Variable T-test df M SD M SD

Question 3a -0.94 13

Question 5 1 .44 6

Question 6 -1.00 4

Question 7a 0.86 10

Question 7b -0.67 12

Question 8a -1.19 13


Note. 1 = no, 2 = sometimes, 3 = yes

Question 3a = Does your child have a toy he/she plays with more often than all the

others?
Question 5 = Does your child grow upset if this toy should go missing or be taken from

him/her?
Question 6 = When your child is upset about something, does he/she seek out that toy for

comfort?
Question 7a = Does your child like to play with his/her peers?
Question 7b = Is there a particular playmate he/she prefers?
Question 8a = Does your child grow upset easily when asked to play with others?


2.29

0.91

2.71

0.49

1.00

0.00

2.82

0.60

2.23

1.01

1.21

0.43


2.57

0.76

2.29

0.64

1.20

0.45

2.59

0.66

2.38

0.68

1.43

0.62


On


61

In addition, the participants' caregivers and educators were asked whether the
children's favourite toys provided them with some sort of sensory stimulation. Most of
the adults (63.6% of the mothers and 81.8% of the educators) said that the toys did
provide sensory stimulation to the participants. Visual and tactile stimulation were cited
most frequently. Auditory stimulation was also mentioned. While certain respondents
believed the children played appropriately with their favourite toys (mothers-40%,
educators-39.3%)), others thought that the participants were repetitious when they played
(mothers-20%, educators-39.3%)).

Although the caregivers and educators mentioned that the participants generally
grew upset if this toy went missing or was taken from them (mothers and educators both
answered between "sometimes" and "yes") , they generally did not tend to seek out that
toy for comfort when they were upset about something else (all of the mothers and most
of the educators responded "no" to this question).

The majority of the respondents perceived the participants as liking to play with
their peers (responses ranged from "sometimes" to "yes" for both educators and
mothers). When asked if there was a particular playmate the children preferred, responses
ranged from "sometimes" to "yes" for both the mothers and the educators.

Those educators (55.9%)) and caregivers (64.7%) who said that the participants
had particular preferred playmates were then asked to answer a series of questions about
these playmates. The following percentages are based on those who answered "yes" to
this question. When asked what was the relationship between the participant and the
preferred playmate, classmates were sometimes mentioned (mothers-27.3%, educators-


62


88.9%). Siblings were also mentioned as preferred playmates (mothers-63.7%, educators-
22.2%). As mothers have more of an opportunity to observe the participants playing with
siblings, this finding is not surprising.

The children's educators and caregivers also were asked to describe how the
participants played with their preferred playmates. Some of the mothers mentioned that
the children would play with cars and trains (25%). In addition, they made reference to
running around and chasing each other (18.8%). Other behaviours included pretend
(18.8%) and rough-and-tumble play (12.5%). The educators listed blocks most
fi"equently as the activity the participants engaged in with their preferred playmates
(15.5%). Like the mothers, the educators mentioned tag and chasing games (12.6%).
Again, this discrepancy is not surprising given that the participants and their playmates
have access to different toys at home than at school.

Although most of the respondents indicated that the children did not grow upset
easily when asked to play with others (responses for both mothers and educators ranged
between "no" and "sometimes"), some of them did grow upset. When asked what kinds
of things the educators and caregivers did to calm down the participants if they did get
upset, most of the respondents said that they would talk to the child (mothers-60%),
educators-55%). The adults believed re-directing the child, removing him or her from the
activity, and /or telling him or her that their behaviour was inappropriate were also
relevant strategies.

The majority of the respondents believed that the participants had the same
opportunities to play as their typically-developing peers (mothers-60%, educators-


63

81.5%), although some were unsure of their response (mothers-6.7%, educators- 18.6%).
One mother believed her child had more opportunities to play than his or her classmates.
When asked why (or why not) the participants did (or did not) have the same
opportunities to play as their peers, explanations included the child does what everyone
else does (mothers-6.3%, educators-41.5%)) and is offered the same resources during play
time (mothers didn't mention, educators-9.6%). At school, the educators made sure that
the participants were allotted the same amount of time to play as their typically-
developing peers (educators-26.7%). In addition, the children with autism were
encouraged by caregivers and educators to participate (mothers-31.3%), educators-
28.1%) and the typically-developing peers were encouraged to interact with the
participants (mothers did not mention, educators- 15.1%) .

Those who did not think that the participants had the same opportunities to play as
their peers (mothers-33.3%, educators-0%) listed reasons such as a lack of early
intervention to target the deficits in social behaviours of children with autism (mothers-
6.3%), educators did not mention), a dearth of community resources to aid in social
learning (mothers-6.3%. educators teachers did not mention), and a general lack of
knowledge about autism spectrum disorders (mothers- 12. 5%, educators did not mention).
In addition, some of the children had to go to various appointments during play time,
such as therapy, focal sessions and special programs (mothers did not mention, educators-
26.7%). Mention was also made that some children with autism may have the same
opportunities to play as their typically-developing peers but they do not always take the
opportunities, either because they lack the desire or the skills to do so (mothers- 18.8%,


64

educators- 12. 6%).

Finally, the adult respondents were asked what they thought their role was in
structuring the participant's play time. Caregivers and educators mentioned their
perceived responsibility to encourage social interaction between the child and his or
her peers (mothers- 12. 5%, educators-69.4%). This was done by being a mediator for the
child with autism. The importance of encouraging a variety of toy choices was also
emphasized (mothers-6.3%, educators- 15.6%). Some of the caregivers and educators
were concerned that the children would continue to play with the same toy constantly
unless they were provided with alternate choices. In addition, a few of the adults believed
their role included insuring the participants' safety (mothers- 12. 5%, educators- 12. 6%).
Reminding the participants of the rules and supervising their activities (mothers-6.3%,
educators-23.5%) was also stated. Furthermore, mention was made of the need to
motivate the participant to play (mothers- 12. 5%, educators-9.3%). By getting the child
started on an activity, some of the adults believed they encouraged the child to participate
(mothers- 18. 8%, educational assistants-5.9%). Not only was motivating the child to play
considered important, but also insuring that the play was appropriate (i.e., not too rough,
making sure the rules were followed, etc., mothers-6.3%, educators-36.1% ). The adults
also made sure that appropriate play was reinforced through praise and motivation to
continue to play properly (mothers- 12. 5%, educators- 12.2%).

Research Question Four: Comparing Observations with
Educator and Caregiver Reports

The final research question addressed in the current study was, "Do observed play


65


behaviours differ from educator and/or caregiver perspectives?" In order to attempt to
respond to this question, I compared the results of my naturalistic observations with the
mothers' and educators answers to the PPBS questionnaire. This was followed by a
correlational analysis of the observed behaviours and the reported behaviours, as
measured by the PPBS and the interview.
Observed vs. Reported Behaviours

Comparing Mean Responses to the PPBS and the POS . Educators' and mothers'
perspectives on participant play behaviours, as measured on the PPBS (i.e., the five-point
Likert scale measuring the frequency of various social and cognitive play categories),
were compared to observed behaviours. In order to insure some degree of similarity
between the categories on the POS (e.g., solitary-functional, parallel-dramatic, etc.) and
the categories on the PPBS (e.g., solitary active, solitary passive, etc.), I regrouped the
PPBS variables so that they more closely matched those variables measured by the POS
(e.g., "Talks to other children during play" = peer conversation). A listing of these
regrouped categories are included in Appendix D.

Questions on the PPBS that reflected a similar behaviour were then added
together and averaged (e.g., peer conversation = "Talks to other children during play" +
"Engages in active conversations with other children during play." 12). The mean score
for each question on the PPBS was calculated for the educators and the mothers. Since
the mothers and educators did not differ significantly in their perspectives as already
measured by the PPBS, their scores were averaged and compared with the observed
frequencies of play behaviours (as indicated in Table 3). Results are summarized in Table


66


10.


Table 10


Mothers' and Educators' Scores on the PPBS. as Compared with Observed Play
Behaviours


Behaviour


Mothers and Educators


Observed


Solitary-exploratory

Solitary-constructive

Solitary-dramatic

Solitary-functional

Onlooker

Group

Group-dramatic

Peer conversation

Unoccupied

Rough-and-tumble


M


Percent of total time


3.62

4.6%

3.78

3.8%

3.12

0.4%

2.99

9.9%

3.09

6.6%

2.70

6.5%

2.45

0.6%

2.46

7.0%

2.50

8.4%

2.24

0.7%


Note. 1 = never, 2 = hardly ever, 3 = sometimes, 4 = often, 5 = very often. The observed
behaviours are based on the percent of total time (i.e., 30 ten-second intervals) that each
behaviour occurred.

According to the mothers and educators, the most frequently occurring play

behaviours were all of the solitary social category, with the cognitive behaviours in the

following descending order: constructive, exploratory, dramatic, and fiinctional. The least

occurring behaviour was rough-and-tumble play. Rankings of the solitary cognitive

behaviours from observations occurred in the following descending order: fiinctional,

exploratory, constructive, dramatic. Based on observations, solitary-dramatic play was

ranked as the least occurring behaviour. Hence, some differences existed between the

rankings of the cognitive play behaviours obtained from observations and how the


67

caregivers and educators ranked them in terms of frequency. Nonetheless, the observed
behaviours more generally resembled educator and caregiver perspectives in that group
behaviours were less frequently noted than solitary behaviours. In summary, there was
agreement between the observations and the perceptions of the participants' caregivers
and educators that solitary play behaviours occurred more frequently than group
behaviours, however, differences existed in the ranking of frequency of cognitive play
behaviours.

Correlating Observed and Reported Behaviours. In order to determine whether the
observed behaviours differed from or resembled those reported by the educators and
mothers, correlational analyses were conducted. The averaged responses on the PPBS for
the mothers and educators were correlated with the average frequency of observed
behaviours. Results are summarized in Table 1 1 .

In general, a positive trend exists between the observed and reported participant
play behaviours. Although some of the correlations were not significantly positive, it is
important to bear in mind the small sample size used in the current study. In addition,
many of the variables used in the correlational analysis were based on a small number of
items. For example, reported solitary-fijnctional and solitary-constructive play were both
based on one PPBS item each. Finally, the negative relationship between observed and
reported solitary-dramatic behaviour indicates some (although not significant)
disagreement between myself and the participants' caregivers and educators.


68


Table 11

Correlations Between Observed and Reported Play Behaviours

Behaviour E

Solitary-exploratory .54*

Solitary-dramatic -.11

Unoccupied .09

Onlooker .41

Group .13

Group-dramatic .39

Peer conversation .84**

Rough-and-tumble . 1 7

Solitary-functional .14

Solitary-constructive .54*

Note, r refers to the correlations between observed behaviours, as measured on the POS,
and reported behaviours, as measured on the PPBS. *£ < .05, **£ < .01.

DISCUSSION

The four research questions that the study attempted to address included: During
free play time at school, at what cognitive and social levels do children with autism
spectrum disorders play? What are teachers' and parents' perceptions of participants'
play behaviours? Are there differences between parent and educator reports of
participants' play abilities? Finally, do observed play behaviours differ from educator
and/or caregiver perspectives? The findings for each research question are now discussed
separately.

Observed Play Behaviours

To address the first question, I observed each participant during free play time at
school. According to Siegel (1996), play centering around one particular toy that the


69


participant lias developed a strong attachment to should be common. In addition,
Siegel (1996), as well as Winzer (1987), believed that many children with autism
spectrum disorders would display an interest in junk items (i.e., string, elastic bands,
empty bottles) that may be stronger than an interest in toys. In addition, it was expected
that toy choice would be based on the potential to provide sensory stimulation. Siegel
(1996) also pointed out that children with autism spectrum disorders have a tendency to
seek out their preferred toys when they are upset about something else. Furthermore, it
was expected that the play taking place would be mainly in the sensorimotor stage in
children with autism (Hughes, 1998; Siegel, 1996; Tilton & Ottinger, 1964). Dramatic
play, games, constructive play and combining different toys during play were likely to be
rare, whereas exploratory behaviour would be the norm for this sample (Siegel, 1996;
Tilton & Ottinger, 1964). Moreover, Hughes (1998) suggests that children with autism
often engage in perseveration, an excessively repetitious form of behavioural expression.

Many of these statements about the cognitive and social play behaviours of
children with autism were not borne out in the observations that I made. First, the
participants in the study rarely (i. e., approximately 2% of the time) exhibited
perseveration during their play. In fact, perseveration was removed as a code on the play
checklist due to its extremely rare occurrence. Thus, the children with autism did not
appear to prefer one toy to any of the others. They might have been repetitive, but they
were able to move on to other activities. Second, not one child engaged in play with junk
items during the observation sessions. This led to the removal of "junk item play" as a
category on the modified POS during the pilot-testing phase of the study. Third, ahhough


70

dramatic play and games were rare, constructive play was the second most frequently
observed cognitive level after functional play during observations.

Play with junk items was very rare in the participants, according to the adult
respondents. The educators reported that this behaviour hardly ever to never occurred,
and the mothers said that it hardly ever occurred. A possible explanation for these
findings is that the children the other researchers referred to were lower functioning than
the children sampled in the study (Siegel, 1996; Tilton & Ottinger, 1964; Winzer, 1987).
In addition, junk items may be less available at school than at home, possibly because the
educators remove them from the classroom.

It is certainly noteworthy that so much functional play occurred within the current
sample. Some previous research indicated that children with autism tend to spend most of
their time in exploratory play or perseveration (Hughes, 1998; Siegel, 1996; Tilton &
Ottinger, 1964), however, Lewis and Boucher (1995) found that children with autism
were equally capable as their peers with learning disabilities and typical development
(younger) at producing ideas for functional play. The current study lends some support to
the work of Lewis and Boucher (1995). This means that the participants in the current
study played with toys according to their functions more frequently than using them as a
source of sensory stimulation.

It is possible that the research described above was conducted using children who
were not as highly functioning as those in the present sample. While Siegel (1996) did
not give specific ages at which children with autism are expected to become attached to a
particular toy to the exclusion of all others or to play with junk items, she did say that


71

both behaviours occur when the children are young. Perhaps she was referring to a
younger group than the participants in the current study. Winzer (1987) stated that
children with autism show a fascination with junk objects at 18 months-2 years, a much
younger age level than the current sample, where the mean age is approximately six
years of age. Siegel (1996) was not specific about how long children with autism remain
in sensory or exploratory play, other than to say that it is for longer than typically-
developing young people. Nor was Winzer (1987), who simply said that little or no group
play emerges in children with autism from the age of two onward. In addition, Siegel
(1996) stated that perseverative behaviours tend to increase with level of delay. Since the
sample in the current study was integrated at approximately the correct level for
participant chronological age, it may be reasonable to conclude that less perseveration
would be expected in my sample. Furthermore, Hughes (1998) pointed out that although
children with autism (with no specific age level cited) have a tendency to perseverate, it
would be unfair to conclude that all children with autism engage in perseveration.

Although the participants in Tilton and Ottinger (1964)'s study resembled those in
the current study in terms of sample size (the current study had 1 7 children, Tilton and
Ottinger' s had 13) and mean age (of approximately five years), the children in Tilton and
Ottinger' s study had not received a diagnosis of an autism spectrum disorder at the time
of the study. In the current study, all but one participant had. In addition, Tilton and
Ottinger' s sample were not testable by standard psychological tests, whereas the majority
of the current sample were able to complete the PPVT-R.

It is unclear whether or not the children reported in these published research


72

studies attended integrated schools, as did the participants in the current study. In
addition, the published research studies were based either on laboratory with no play-
mates or strangers, or on case studies and clinical experience. It is likely that there are
differences between the play behaviours that could be observed in a child who is in a
novel situation surrounded by strangers and those observed in a child who is in a situation
to which he or she is accustomed, surrounded by familiar peers. For example, it is
possible that more anxious behaviours occur when the child is unsure of his or her
surroundings. In addition, the case study method is less generalizable to other children
(Huck & Sandler, 1979).

According to my observations, the most frequently occurring play behaviour in
the sample was parallel-functional play (occurring approximately 13% of the time).
Functional play was the most frequently occurring cognitive type of play. This finding
lends some support to the work of Lewis and Boucher (1995), who concluded that
children with autism did not differ from other children in their ability to produce
functional play. However, I did not observe as many instances of symbolic play as
Lewis and Boucher, who contended that children with autism were equally capable of
producing symbolic play ideas. Dramatic play (i.e., symbolic play, make believe) was
rarely observed in the participants of the current study. This may have been due to the
fact that Lewis and Boucher (1995) used participants with a higher mean chronological
(1 1 years, 10 months) and mental (5 years, 9 months) age than the current sample. In this
sense, the current research more closely resembled that of Rogers (1988), who found that
children with autism spend more of their time in less mature forms of play (i.e., less


73

games and dramatic play, more exploratory and functional play).

Parallel and solitary play were the most frequently observed social play
behaviours. In comparison, group play was quite rare. The remainder of the time was
spent by the participants in interactions with adults, peer conversations, out of the room,
crying, or anxious, aggressive, transitional, unoccupied and onlooker behaviours. These
findings are similar to those of Lord (1984), who concluded in her ovm research that
children with autism spend more time than typically-developing children playing alone
when presented with the option of playing with other children.

Relationship to typically-developing children of the same age . The participants in
the current study had a mean verbal mental age of 4.86 years and a mean chronological
age of 6.21 years. Most of the young people were in junior kindergarten, senior
kindergarten or Grade 1 at the time of the study. According to previous research, by the
time they are in kindergarten, typically-developing children should be engaging in more
parallel-constructive, parallel-dramatic and group-dramatic play and less unoccupied and
onlooker behaviours, solitary-functional and parallel-functional play than they did at
preschool age (Rubin et al., 1978).

The current sample of children with autism appeared to be playing more at the
preschool level than at the school-age level. Most of their time was spent in either
parallel- or solitary-functional play. Onlooker and unoccupied behaviours occurred more
frequently than did most of the other play behaviours. Although parallel-constructive play
was fairly frequent, parallel- and group-dramatic behaviours were quite rare. The concept
of pretense emerges between three to six years of age in typically-developing children


74

(Rubin et al., 1983). It was rarely witnessed in the current sample of participants of
approximately 6 years of age. Since the children with autism were playing at less mature
social and cognitive levels than their typically-developing peers, it is possible that the
participants in the present study experienced a delay in their development .

Parent and Educator Perspectives

In order to address the second research question, the participants' caregivers and
educators were asked to provide their perceptions of the participants' play behaviours.
This was achieved through having them complete the Preschool Play Behaviour Scale
(Coplan & Rubin, 1998) and through interviewing the adult respondents. In general, the
findings here also do not support those of previous researchers.

Although most of the adults did believe that the participants had one particular toy
that they preferred to all the others, and that that toy provided the children with visual and
tactile stimulation, rarely was it the case that the participant sought out that toy for
comfort when he or she was upset about something. This may be due in part to the fact
that, while there was a preferred toy, the preference tended to change from week to week.
For example, while the child might have preferred a toy truck one week, the next week it
might have been a doll. In addition, less than half of the participants' caregivers and
educators believed that the children engaged in repetitive play. The others believed that
the children played appropriately with their favourite toys. Thus, although the
participants in this study might have preferred one certain toy, it does not seem that the
children were attached to it. In this situation, attachment refers to seeking comfort from
an object, not unlike a typically-developing child of a similar chronological age going to


75

a primary caregiver when he or she is unhappy (Siegel, 1996).

The adult respondents also were asked about the social behaviours the participants
engaged in at home or at school. Many of the educators and caregivers believed that the
children liked playing with their peers and that there was a particular playmate that each
participant preferred. While Lord (1984) believed that children with autism had
difficulties engaging in age-appropriate play with peers, the findings of the current
research seemed to indicate the contrary. Perhaps this is because Lord based her
conclusions on laboratory observations of interactions with strangers, whereas the present
study was based on both naturalistic observations of interactions with familiar peers and
educator and parent interviews. It may have been the case that the educators, caregivers
and I observed more social play because the children with autism were granted the
opportunity to play with children they see on a daily basis, as opposed to Lord's study,
where the children with autism were meeting their playmates for the first time.
Nonetheless, a fair number of respondents believed that the children would grow upset at
least sometimes when asked to play with other children.

Comparing Parent and Educator Perspectives

The third research question I wished to answer was: "Are there differences
between parent and educator reports of participants' play abilities?" Comparing the
responses of the educators and the mothers on the PPBS and the interview aided in
responding to this inquiry. The comparisons yielded no significant differences between
the mothers and the educators in terms of their responses to the PPBS and the interview
questions. It appears that the play behaviours of children with autism at home do not


76

differ from the types of play they engage in at school. This might indicate consistency
between the perspectives of the educators and the caregivers on the participants' play
behaviours.

Comparing Adult Perspectives With Observed Behaviours

Finally, I wished to compare caregiver and educator perception of participant play
behaviours with observed participant play activities. This was accomplished by
correlating the responses made by the educators and caregivers to the PPBS with the
observed behaviours recorded on the POS. A positive trend was found (i.e., increased
reporting of a certain behaviour was related to increased observations of that behaviour).
In general, the reported and observed play behaviours of the children with autism were
similar. Not all of the correlations were significant, and one was negative but
nonsignificant. Possible explanations for this, besides the aforementioned small sample
size and lack of items making up the subscales, are discussed below.

Methodological Considerations

In this section, I will attempt to account for the discrepancies between teacher and
maternal reports and my observations of the participants' play behaviours. In addition, I
will discuss general methodological considerations that apply to the entire research
project.

A Comparison Between Maternal and Teacher Reports
and Observed Play Behaviours

There are a few possible explanations as to why my observations differed from
the reports of the participants' mothers and educators. The differences could be due to


77


problems inherent in the observational method of research. For example, the children
might have been aware that they were being observed and they may have behaved
differently than if the observers were not present. Moreover, the observational method is
very difficult to replicate. Even with operationalizations of the behaviours on the
checklist, there is always going to be an element of subjectivity involved in the coding of
observed behaviours (Huck & Sandler, 1979). There may have also been subjectivity
involved in the adults' responses to the PPBS. Furthermore, there may have been
problems with the coding system itself The POS requires the coder to select a code from
a list of several categories. This sometimes made it difficult to choose which code to use
for a given behaviour. However, this probably did not affect the results a great deal,
considering that the inter-rater reliability for the POS was good.

General Methodological Considerations

There are a few caveats that should be acknowledged about interpreting the
results of this study. Within any group of children, individual differences due to such
factors as ability level, personality and temperament may impact on the results. For
example, a child with autism who has a withdrawn personality might be less willing to
engage in group play than a more extroverted child. Even if these two children both have
autism, their personalities are still going to differ. Add to this the varying severity of the
individual child's autism spectrum disorder and it is not surprising that there is so much
variability among children with autism in general (Siegel, 1996).

Furthermore, individual differences exist within the adults who participated in the
study. Perhaps some of the educators might have encouraged some children to play more


78

than others. For example, some, but not all, of the adults believed it was their
responsibility to encourage social interaction between the participants and their typically-
developing peers during free-play time. Others thought they should be involved in the toy
choices made by the children. Motivation, encouragement and praise for appropriate play
was also mentioned by a few of the educators and caregivers.

Alternatively, there might be something about the adults who agreed to allow their
child or student to participate that differed from the those who refused to volunteer for the
research, affecting the generalisability of this study.

The possibility of social desirability should also be considered, particularly
among the adult participants. Being aware that an observer is present in the classroom
might have caused some of the educators to get more involved in the participants' play
than they might normally have done. Interview and questionnaire responses also might
have been influenced by a social desirability bias.

Finally, it is important to remember that the current study was different from
those mentioned in the literature review because it was a naturalistic observation, as
opposed to a laboratory study. In addition, most of the children in the current sample
were diagnosed with an autism spectrum disorder at approximately 3 years of age. Due to
their early diagnoses, some of them have been attending special programs and all of them
have had an educational assistant present in the classroom. This means that they might
have experienced some intervention geared at increasing play skills already.

Summary

The first and main research question addressed in the current study was, "During


79

free play time at school, at what cognitive and social levels do children with autism
play?" The participants with autism in this study engaged most frequently in parallel and
solitary play. This indicated a preference to play alone or near, as opposed to with, other
children. Group play was rare in comparison. Functional, exploratory and constructive
play were more likely to occur within this group than were dramatic play or games. The
participants almost never engaged in perseveration when playing , nor did they use junk
items as playthings. Sometimes, they would decline to play, preferring instead to
spectate, wander about aimlessly or stare into space.

Although the children sometimes had a favourite toy that they would rather play
with than any of the others, the play itself was not considered perseverative. Furthermore,
it was very unlikely that the participants would seek out that toy for comfort when they
were upset about something else. Functional toys (such as train sets) and constructive
toys (such as Lego and blocks) were mentioned most frequently as examples of the
participants' favourite playthings. These toys might have provided the child with sensory
stimulation, however, play was usually appropriate rather than perseverative.

The participants with autism were not likely to grow upset easily when asked to
play with other children. In fact, they seemed to enjoy playing with their peers. Some of
them even preferred a certain person as their playmate.

According to published research literature, these children are playing at a
preschool level. Since they are approximately six years of age, they should be engaging
in more group and dramatic play and less solitary and functional play than was actually
witnessed. This does not appear to be due to a lack of opportunities for the children with


80

autism to play. The following section provides some suggestions for fiiture investigation
which may help to isolate the reasons why children with autism spectrum disorders do
not play at the same cognitive and social levels as their typically-developing peers.

Suggestions for Future Investigation
This study provides baseline data for future research on the play behaviours of
children with autism spectrum disorders. It would be useful to expand on the current
research by exploring related topics. Now that there is a general idea of what types of
social and cognitive play children with autism engage in, it would be interesting to
compare these findings with the play behaviours of age-matched typically-developing
children and children with other developmental disabilities. For instance, it would appear
that children with Down Syndrome are often included as a comparison group in play
studies because researchers would like to see if and how the diagnostic groups differ.
Children with Down Syndrome serve as a good control group because, although they may
have learning difficulties, they do not have the communication, imagination and
socialization problems associated with autism (Libby et al., 1997). Without a typically-
developing control group and another diagnostic group for comparison purposes, it would
be difficult to conclude that the differences in play behaviour that may be found are
specific to autism. Moreover, it would be interesting to compare play behaviours
between children who have different diagnoses under the autism spectrum. Although
children with the autism diagnosis share similar behaviours, the reasons why they were
diagnosed with autism are usually quite different from child to child. Comparisons
between groups of children with different types of autism might be beneficial in terms of


C*J<


81

identifying any play deficits unique to particular groups and attempting to address these
deficits in intervention programs.

In the current study, a comparison group was not included due to time constraints.
In order to conduct a detailed examination of the participants, seven visits were required
for each one. Thus, the data from the present study might be considered as a first
step towards future research that might include a control group of typically-developing
peers.

It also may be useful to examine the effects of scaffolding behaviours of
typically-developing children on the play behaviours of children with autism spectrum
disorders and other developmental disabilities. Perhaps play with typically-developing
peers might enhance the play of children with developmental disabilities in that these
children might engage in more sophisticated cognitive and social levels of play. Some
research exists that supports this idea (Lord & Hopkins, 1986; McHale, 1983; Roeyers,
1996).

There also are some suggestions regarding replicating the current study. First,
although I tried to include a second primary caregiver (i.e., fathers, grandparents, etc.) in
the study, it was not always possible for the second caregiver to attend the home visit.
Therefore, not enough of them were available to make it possible to analyze their data.
Future research might benefit from including the responses of second primary caregivers
to see if they differ from the perspectives of the first primary caregivers and the
educators. If these perspectives do differ, perhaps the second caregivers have noticed
something that the other respondents might have missed. It might also be the case that


11^


82

children play differently with certain caregivers than they do with others (e.g., maybe
they engage in more rough-and-tumble play with their fathers than their mothers, Jafife,
1997).

Second, it would be useful to include a question on the PPBS regarding how
frequently the participants engaged in parallel play. Since I observed parallel play as the
most frequently occurring social play behaviour, it would be interesting to see whether
the educators and caregivers concur with this opinion.

Third, a larger sample consisting of an equal number of males and females, more
varying racial backgrounds and enough representatives from each grade level to make
comparisons possible would have been beneficial. Developmental trends might have been
observed in the play abilities between age groups. Gender and cultural differences would
also have been valuable to report.

Finally, it might be useful to conduct observations in a manner in which the
participants are less aware of the presence of an observer. Perhaps observers could come
in a few times before actually conducting observations. When it is actually time to start
the study, the presence of the observers would no longer be novel. Thus, the participants
would be less likely to be uncomfortable playing with the observers around.
Alternatively, the presence of a camera in the classroom (if installed a few days in
advance and only turned on during the study) might be less intrusive than a strange
person.

Conclusions and Practical Implications

As outlined in the results of this study, no differences existed between maternal


83


and educator ratings and researcher observations of the play behaviours of children with
autism. However, the participants were playing at a lower developmental level than their
typically-developing peers. Overall, it is clear that children with autism engage most
often in play alone but frequently near others. While school age typically-developing
children were expected to exhibit more parallel-constructive, parallel-dramatic and
group-dramatic play and less unoccupied and onlooker behaviours, solitary-functional
and parallel-functional play than they did at preschool age, the children with autism in the
current study played more like typically-developing preschoolers. The most frequently
occurring play behaviours for the children with autism included parallel-functional and
solitary-functional play and unoccupied behaviours. The least frequently occurring play
behaviours included parallel- and group-dramatic activities.

A few practical implications of the study are worth considering. A need exists to
promote more sophisticated levels of play in children with autism spectrum disorders.
Results of the current study should definitely be shared with all adults who play a vital
role in the child with autism' s development. Discussion becomes important in order to
ensure the child with autism is given every opportunity to play and develop as his or her
typically-developing peers. In addition, every effort should be made to provide the child
with autism with as many different play resources as possible. By giving these children as
many chances and materials as their typically-developing peers, they are also granted the
opportunity to develop socially, intellectually and emotionally.


84


REFERENCES

American Psychiatric Association. (1994). Diagnostic and statistical manual of
mental disorders (4111 ed.). Washington, DC: Author.

Aylott, J. (2000). Understanding children with autism: Exploding the myths.
British Journal of Nursing . 9 02 ). 779-784.

Barnard, K. E., Bee, H. L., & Hammond, M. A. (1984). Home environment and
cognitive development in a healthy, low-risk sample: The Seattle study. In A. W.
Gottfried (Ed.), Home environment and early cognitive development: Longitudinal
research . New York: Academic Press.

Baron-Cohen, S. (1993). Theory of mind and autism: A fifteen-year review. In
Baron-Cohen, S., Tager-Flusber, H., & Cohen, D. J. (Eds.), Understanding other minds:
Perspectives fi'om autism (2"^ ed.). Oxford: Oxford University Press.

Baron-Cohen, S., & Bolton, P. (1993). Autism: The facts . New York: Oxford
University Press.

Baron-Cohen, S., Leslie, A.M., & Frith, U. (1985). Does the autistic child have a
theory of mind? Cognition. 21 . 37-46.

Bednersh, F., & Peck, C. A. (1986). Assessing social environments: Effects of
peer characteristics on the social behaviour of children with severe handicaps. Child
Study Journal. 16 (4), 315-329.

Bettelheim, B. (1967). The empty fortress: Infantile autism and the birth of the
self New York: Free Press.


85


Bjorklund, D. F., & Brown, R. D. (1998). Physical play and cognitive
development: Integrating activity, cognition and education. Child Development , 69 (3),
604-606.

Black, M., Freeman, B. J., & Montgomery, J. (1975). Systematic observation of
play behaviour in autistic children. Journal of Autism and Childhood Schizophrenia . 5
(4), 363-371.

Bradley, R. H. (1986). Play materials and intellectual development. In A. W.
Gottfried & C. C. Brov^n (Eds.), Play interactions: The contributions of play materials
and parental involvement to children's development (pp. 227-252). Lexington, MA:
Heath.

Bradley, R. H. & Caldwell, B. M. (1984). 174 children: A study of the
relationship between home environment and cognitive development during the first five
years. In A. W. Gottfiied (Ed.), Home environment and early cognitive development:
Longitudinal research . New York: Academic Press.

Byers, J. (1998). The biology of human play. Child Development . 69 (3), 599-
600.

Charman, T. (1997). The relationship between joint attention and pretend play in
autism. Development and Psychopathology . 9 (1), 1-16.

Charman, T., & Baron-Cohen, S. (1997). Brief report: Prompted pretend play in
autism. Journal of Autism and Developmental Disorders. 27 (3), 325-332.


86


Charman, T., Swettenham, J. S., Baron-Cohen, S., Cox, A., Baird, G. & Drew, A.
(1997). Infants with autism: An investigation of empathy, pretend play, joint attention,
and imitation. Developmental Psychology . 33 (5), 781-789.

Coplan, R. J., & Rubin, K. H. (1998). Exploring and assessing nonsocial play in
the preschool: The development and validation of the Preschool Play Behaviour Scale.
Social Development , 7. (1), 72-91.

Coplan, R. J., Rubin, K. H., Fox, N. A, Calkins, S. D., & Stewart, S. L. (1994).
Being alone, playing alone and acting alone: Distinguishing among reticence and passive
and active solitude in young children. Child Development, 65. 129-137.

DeKlyen, M., & Odom, S. L. (1989). Activity structure and social interactions
with peers in developmentally integrated play groups. Journal of Early Intervention . 13
(4), 342-352.

Freud, S. (1959). Creative writers and daydreaming. In J. Strackey (Ed.), The
standard edition of the complete works of Sigmund Freud (Vol. IX). London: Hogarth.

Geneva Centre (1997). What is autism/P.D.D?: An Introductory Workshop .
Training workshop presented to the Autism Society of Ontario, North York Chapter,
Toronto, Ontario.

Harter, S. (1983). Developmental perspectives on the self-system. In P. H.
Mussen (Ed.), Handbook of child psychology (4^ ed.. Vol. 4, pp. 275-385). New York:
Wiley.

Hartley, R. E., Frank, L. K., & Goldenson, R. M. (1952). Understanding
children's play . New York: Columbia University Press.


87

Hill, P. M., & McCune-Nicholich, L. (1981). Pretend play and patterns of
cognition in Down Syndrome. Child Development . 52, 6 1 1 -6 1 7.

Hollenbeck, A. R. Problems of reliability in observational research. In G. Sackelt
(Ed.), Observing behaviour Vol. 2: Data collection and analyses methods. Baltimore:
University Park Press.

Huck, S. W., & Sandler, H. M. (1979). Rival hypotheses: Alternative
interpretations of data based conclusions. NY: HarperCollins.

Hughes, F.P. (1998). Play in special populations. In O. N. Saracho and B. Spodek,
(Eds.), Muhiple perspectives on play in early childhood education . Albany: State
University of New York Press.

Hughes, F. P. (1999). Children, play, and development Ord ed.) . Toronto: Allyn
and Bacon.

Hymel, S., Rubin, K. H., Rowden, L., & LeMare, L. (1990). Children's peer
relationships: Longitudinal prediction of internalizing and externalizing problems from
middle to late childhood. Child Development . 61, 2004-2021.

Jafife, M. L. (1997). Understanding Parenting (2"^ ed.) . Toronto: Allyn and Bacon.

Jarrold, C, Boucher, J., & Smith, P. K. (1996). Generativity deficits in pretend
play in autism. British Journal of Developmental Psychology . 14, 275-300.

Keppel, G. (1991). Design and analysis: A researcher's handbook (3*^"^ ed.).
Englewood Cliffs, NJ: Prentice Hall.

LeMare, L. J., Rubin, K. H. (1987). Perspective taking and peer interaction:
Structural and developmental analyses. Child Development. 58. 306-315.


r^


88


Lewis, v., & Boucher, J. (1995). Generativity in the play of young people with
autism. Journal of Autism and Developmental Disorders . 25 (2), 105-121.

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(4), 365-383.

Libby, S., Powell, S., Messer, D., & Jordan, R. (1998). Spontaneous play in
children with autism: A reappraisal. Journal of Autism and Developmental Disorders , 28
(6), 487-497.

Lord, C. (1984). Development of peer relations in children with autism. In F.
Morrison, C. Lord, & D. Keating (Eds.), Applied developmental psychology (pp. 165-
229). New York: Academic Press.

Lord, C, & Hopkins, J. M. (1986). The social behaviour of autistic children with
younger and same-age nonhandicapped peers. Journal of Autism and Developmental
Disorders . 16(3) . 249-262.

Malone, D. M., & Stoneman, Z. (1990). Cognitive play of mentally retarded
preschoolers: Observations in the home and school. American Journal of Mental
Retardation . 94 (6), 475-487.

Maurice, C. (1993). Rescuing my daughter. McCalFs . June, 81-84,156.

McCabe, J. R., Jenkins, J. R., Mills, P. E., & Dale, P. (1996). Effects of play
group variables on language use by preschool children with disabilities. Journal of Early
Intervention . 20 (4), 329-340.


89

McHale, S. M. (1983). Social interactions of autistic and nonhandicapped children
during free play. American Journal of Orthopsychiatry . 53 (W 81-91.

Morgan, S. B., Cutrer, P. S. (1989). Do autistic children differ from retarded and
normal children in Piagetian sensorimotor fimctioning? Journal of Child Psychology and
Psychiatry . 30 (6), 857-864.

Nunnally, J. C. (1978). Psychometric theory (2"^^ ed.). Toronto: McGraw-Hill.

Piaget, J. (1962). Play, dreams and imitation in childhood . New York: Norton.

Reber, A. S. (1985). Play. In The penguin dictionary of psychology (pp. 552-553).
Markham, Ontario: Penguin Books Canada Limited.

Riguet, C. B., Taylor, N. D., Benaroya, S., & Klein, L.S. (1981). Symbolic play in

autistic, Down's, and normal children of equivalent verbal mental age. Journal of Autism
and

Developmental Disorders . 1 1 (4) . 439-448.

Roeyers, H. (1996). The influence of nonhandicapped peers on the social
interactions of children with a pervasive developmental disorder. Journal of Autism and
Developmental Disorders , 26 (3) . 303-319.

Rogers, D. L. (1985). Relationships between block play and the social
development of children. Early Child Development and Care, 20 , 245-261.

Rogers, S. J. (1988). Cognitive characteristics of handicapped children's play: A
review. Journal of the Division for Early Childhood , 12 (2) , 161-168.


90


Rubin, K. H., & Coplan, R. J. (1998). Social and nonsocial play in childhood: An
individual differences perspective. In Saracho, O. N., & Spodek, B. (Eds.), Multiple
perspectives on play in early childhood education . Albany: State University of New York
Press.

Rubin, K. H., Fein, G. G., & Vandenberg, B. V. (1983). Play. In P. H. Mussen
(Series Ed.). Handbook of child psychology: Vol.4. Socialization, personality, and social
development (4* ed., pp. 693-774). New York: Wiley.

Rubin, K. H., Maoini, T. L., & Hornung, M. (1976). Free play behaviors in
middle-and lower-class preschoolers: Parten and Piaget Revisited. Child Development,
47,414-419.

Rubin, K. H., Watson, K. S., & Jambor, T. W. (1978). Free-play behaviours in
preschool and kindergarten children. Child Development, 49. 534-536.

Rutter, M. (1973). The assessment and treatment of pre-school autistic children.
Early Childhood Development and Care . 3, 13-29.

Saltz, E., Dixon, D., & Johnson, J. (1977). Training disadvantaged preschoolers
on various fantasy activities: Effects on cognitive fiinctioning and impulse control. Child
Development. 48. 367-380.

Siegel, B. (1996). Social development in autism and PDD. In The world of the
autistic child . New York: Oxford University Press.

Silvern, S. B., Williamson, P. A., & Waters, B. (1982). Play as a mediator of
comprehension: An ahernative to play training. Educational Research Quarterly, 7 , 16-
21.


.^^â– '1: /


91

Smilansky, S. (1968). The effects of sociodramatic play on disadvantaged
children. Toronto: John Wiley & Sons, Inc.

Sullivan, H. S. (1953). The interpersonal theory of psychiatry . New York: W. W.
Norton & Company Inc.

Tager-Flusberg, H., Joseph, R., & Folstein, S. (2001). Current directions in
research on autism. Mental Retardation and Developmental Disabilities Research
Review . L(l), 21-29.

Tilton, J. R., & Ottinger, D. R. (1964). Comparison of the toy play behaviour of
autistic, retarded, and normal children. Psychological Reports . 15, 967-975.

Werner, P. H., & Burton, E. C. (1979). Learning through movement: Teaching
cognitive content through physical activities . St. Louis: Mosby.

Wing, L., & Attwood, A. (1987). Syndromes of autism and atypical development.
In C. Cohen & A. Donnellan (Eds.), Handbook of autism and pervasive developmental
disorders . New York: Wiley.

Winzer, M. (1990). Children with pervasive developmental disorders. In M.
Winzer. (1990). Children with exceptionalities: A Canadian perspective (2"^^ ed., pp.
612-648). Scarborough, Ont.: Prentice-Hall Canada, Inc.

Wolfberg, P. (1999). Play and Imagination in Children with Autism . NY, New
York: Teachers College Press.

Wolfgang, C. H., & Stakenas, R. G. (1985). An exploration of toy content of
preschool children's home environments as a predictor of cognitive development. Early
Childhood Development and Care . 19, 291-307.


92
Appendix Al

BROCK UNIVERSITY. DEPARTMENT OF PSYCHOLOGY

Informed Parental Consent Letter

Date of Ethics Approval: August 23, 2000

Title of Study: "A Naturalistic Observation of the Play Behaviour of Children with
Autism Spectrum Disorders"

Researchers: Elizabeth Holmes, Graduate Student, Psychology Department, Brock
University, 688-5550 ext. 4419, ext. 4519

Teena Willoughby, Thesis Supervisor, Psychology Department, Brock
University, 688-5550 ext. 4067

To the parents of potential research participants:

Your child has been invited to participate in a research project examining the play
behaviour of children with autism spectrum disorders (i.e., Rett's Disorder, Childhood
Disintegrative Disorder, Asperger's Disorder and Pervasive Developmental Disorder-Not
Otherwise Specified). The researchers listed above are conducting this study at Brock
University to fulfill the requirements of an MA Thesis.

Purpose of the study . The purpose of this study is to determine what types of play
behaviour children with autism engage in during free play time at school. This research is
relevant to conduct for a number of reasons. First, this information will help determine
what type of intervention would work best for improving the cognitive level and amount
of play engaged in by children with autism.

Second, there is a general dearth of research in this area. Although many researchers have
examined the differences in play behaviours between children with autism and children
belonging to other diagnostic groups, the participants were all examined on an individual
basis. Virtually no work has been done on how these children interact with each other in
groups, or on how they behave in a naturalistic setting. This research is important to carry
out in order to compare children with autism and typical development in terms of types of
toys chosen, kinds of play behaviour engaged in, and whether or not any social
interaction takes place between the diagnostic groups. It is possible that valuable
information has been missed due to this oversight.

Procedure . The study will take place mainly in your child's school. Your child will be
observed as he/she engages in free play in his/her regular classroom. Your child will be
observed a minimum of five times. Each session will last for approximately the duration


93


of the class' free play time, with the exception of the first session, when I will be present
for an additional ten minutes in order to get your child accustomed to their presence. The
primary researcher will record the child's cognitive and social level of
play, toy choice and any other behaviour observed in the child with autism for the
duration of the play period. An additional researcher will be present until inter-rater
reliability is established. I will not interfere with your child's activities. They will be
present to observe only.

We would also appreciate it if you and your child's teacher (and Educational Assistant if
applicable) would complete a series of questions about your child's typical play
behaviours in the home and classroom (see attached for a copy of the questions). In order
to make the experience more convenient for you, the researcher can meet you at your
home to conduct the interview. If you would rather the interview is conducted at the
school, arrangements can be made to accommodate you there. We would also like to
meet with your child to conduct a vocabulary ability test, with either yourself or the
child's educator present in the room. If your child grows upset by the researcher's
presence, you will be asked to remind him/her why the test is taking place and to signal
the researcher when it is OK to continue testing. If your child becomes upset again,
testing will be stopped immediately. You can decide at this point if you would allow the
researcher to return at a later date. The visit to your home will take between an hour and
an hour-and-a-half

All of the information that we record will be kept completely confidential and only group
data will be reported. That means that your child's name will not appear on any of the
records that we keep. Only the researchers named above and their assistants will have
access to the data. You and your child will be free to withdraw your participation at any
time without penalty.

We hope that you will be willing to participate in our research project and we look
forward to sharing our findings with you at the end of this study. We have attached a
consent form for you to let us know whether you would be willing to let your child
participate in this study.

The study described above has been reviewed and received ethics approval through the
Brock University Research Ethics Board. If you have any concerns or questions about
yours or your child's involvement in the study, you may contact the Director of the
Office of Research Services at 905-688-5550, ext. 4315.


94

Appendix A2

BROCK UNIVERSITY, DEPARTMENT OF PSYCHOLOGY

Informed Educator Consent Letter — Teacher & EA


Date of Ethics Approval: August 23, 2000

Title of Study: "A Naturalistic Observation of the Play Behaviour of Children with
Autism Spectrum Disorders"

Researchers: Elizabeth Holmes, Graduate Student, Psychology Department, Brock
University, 688-5550 ext. 4419, ext. 4519

Teena Willoughby, Thesis Supervisor, Psychology Department, Brock
University, 688-5550 ext. 4067

To the educators of potential research participants:

Your student has been invited to participate in a research project examining the play
behaviour of children with autism spectrum disorders (i.e., Rett's Disorder, Childhood
Disintegrative Disorder, Asperger's Disorder and Pervasive Developmental Disorder-Not
Otherwise Specified). The researchers listed above are conducting this study at Brock
University to fulfill the requirements of an MA Thesis.

Purpose of the study . The purpose of this study is to determine what types of play
behaviour children with autism engage in during free play time at school. This research is
relevant to conduct for a number of reasons. First, this information will help determine
what type of intervention would work best for improving the cognitive level and amount
of play engaged in by children with autism.

Second, there is a general dearth of research in this area. Although many researchers have
examined the differences in play behaviours between children with autism and children
belonging to other diagnostic groups, the participants were all examined on an individual
basis. Virtually no work has been done on how these children interact with each other in
groups, or on how they behave in a naturalistic setting. This research is important to carry
out in order to compare children with autism and typical development in terms of types of
toys chosen, kinds of play behaviour engaged in, and whether or not any social
interaction takes place between the diagnostic groups. It is possible that valuable
information has been missed due to this oversight.

Procedure . The study will take place in your school. You will be asked to make your
student aware of the researchers each time they observe a session. For the first session.


Mj


95


you will be asked to introduce the researchers to your class. Your student will be
observed as he/she engages in free play in his/her regular classroom. Your student will be
observed a minimum of five times. Each session will last for approximately the length of
the class' free play time, with the exception of the first sessiqn, when the researchers will
be present for an additional ten minutes in order to get your student accustomed to their
presence. If you feel your student needs more time, I would appreciate it if you let me
know and I will wait longer before beginning the observations. The primary researcher
will record the child's cognitive and social level of play, toy choice and any other
behaviour observed in the child with autism for the duration of the play period. An
additional researcher will be present until inter-rater reliability is established. I will not
interfere with your student's activities. They will be present to observe only.

Should your student become agitated by the presence of the researchers at any point in
the study, please ask the researchers to leave the room. It would be greatly appreciated if
you could remind the child who we are and what our purpose is once he/she has calmed
down. Once he/she agrees that it is OK for us to return, we will resume our observations.
If he/she continues to feel uncomfortable, we will end the observations immediately.

We would appreciate it if you and your student's Educational Assistant (if applicable), as
well as your student's parents, would complete a series of questions about your student's
typical play behaviours in the classroom and home (see attached for a copy of the
questions, *Note: You are not required to fill those questions in. They are merely an
example of what you can anticipate being asked at the interview). This will take about an
hour. We would also like to meet with your student to conduct a vocabulary ability test,
with either yourself or the child's parent(s) present in the room. If your student grows
upset by the researcher's presence, you will be asked to remind him/her why the test is
taking place and to signal the researcher when it is OK to continue testing. If your student
becomes upset again, testing will be stopped immediately. You can decide at this point if
you would allow the researcher to return at a later date.

All of the information that we record will be kept completely confidential and only group
data will be reported. That means that your student's name will not appear on any of the
records that we keep. Only the researchers named above and their assistants will have
access to the data. You and your student will be free to withdraw your participation at any
time without penalty.

We hope that you will be willing to participate in our research project and we look
forward to sharing our findings with you at the end of this study. We have attached a
consent form for you to let us know whether you will be willing to participate in this
study. If you are not interested in participating, please write "not interested" on the form,
do not sign it, but do return it.


96


The study described above has been reviewed and received ethics approval through the
Brock University Research Ethics Board. If you have any concerns or questions about
yours or your student's involvement in the study, you may contact the Director of the
Office of Research Services at 905-688-5550, ext. 4315.


97
Appendix Bl

BROCK UNIVERSITY, DEPARTMENT OF PSYCHOLOGY
Informed Parental Consent Form

Title of Study: "A Naturalistic Observation of the Play Behaviour of Children with Autism Spectrum
Disorders"

Researchers: Teena Willoughby and Elizabeth Holmes

Name of child participating in study (please print):


I understand that the purpose of the study is to determine the nature of the play behaviours of
children with autism while at school. I understand that this study, in which I have agreed to allow my
child to participate, will involve a researcher observing and recording his/her activities during free
play time in the classroom. The researcher(s) will be present for at least five (5) free play sessions.

My involvement in this research as a parent will include the completion of some written
questionnaires about my child's demographic background and his/her play behaviour at home. The
researcher will come to my home to do this. At this time, my child will also complete a vocabulary
ability test with the researcher. I will also be verbally interviewed regarding my opinions of my
child's ability to engage in certain activities. This will take about an hour to an hour-and-a-half.

I also give permission to the researchers to interview my child's teacher and/or Educational Assistant
regarding my child's ability to engage in different play activities. The interview format is similar to
what I will be given.

I have read and understood the information relevant to the study in the Informed Parental Consent
Letter.

I understand that participation in this study is voluntary and that my child and I may withdraw from
the study at any time and for any reason without penalty. There will be no financial compensation for
participation in this study.

I understand that there is no obligation to answer any question/participate in any aspect of this
project that I consider invasive, offensive or inappropriate.

I understand that all personal data will be kept strictly confidential and that all information will be
coded so that my name and my child's name are not associated with my answers. This consent form
will be kept separate from the questionnaires.

I understand that only the researchers named above and their lab assistants will have access to the
data.

Parent/Guardian Name (please print)


Parent/Guardian Signature Date


Best way to be contacted about possible interview date & where you would like to meet (i.e., phone number
and good time to call, e-mail address etc., would you like to meet at your home, the school,
etc.)


98


Please sign and date this form and return it to your child's teacher. This study has been reviewed and
approved by the Brock Research Ethics Board. (File # 99-342)

If you have any questions or concerns about yours or your child's participation in the study, you may
contact Elizabeth Holmes at (905) 688-5550, ext. 4419, ext. 4519 or Professor Teena Willoughby at (905)
688-5550, ext. 4067. Feedback about the use of the data collected will be available during the month of
February, 2001. A written explanation will be provided for you upon request. Thank you for your help!
Please take one copy of this form with you for further reference .

:|e ]|c :|c :|c :|c % % )|c

I have fully explained the procedures of this study to the above volunteer.

Researcher Signature Date


99
Appendix B2

BROCK UNIVERSITY, DEPARTMENT OF PSYCHOLOGY

Informed Educator Consent Form

Title of Study: "A Naturalistic Observation of the Play Behaviour of Children with Autism Spectrum
Disorders"

Researchers: Teena Willoughby and Elizabeth Holmes

Name of child participating in study (please print);


I understand that the purpose of the study is to determine the nature of the play behaviours of
children with autism while at school. I understand that this study will involve a researcher observing
and recording my student's activities during free play time in the classroom. The researcher(s) will
be present for at least five (5) free play sessions. I understand that my student's parents have
consented to his/her participation in the study.

My involvement in this research as an educator will include the completion of some written
questionnaires about my student's play behaviour at school. I will also be verbally interviewed
regarding my opinions of my student's ability to engage in certain activities. At this time, my student
will also complete a vocabulary ability test with the researcher. This should take about an hour.

I have read and understood the information relevant to the study in the Informed Educator Consent
Letter.

I understand that participation in this study is voluntary and that my student and I may withdraw
from the study at any time and for any reason without penalty. There is no financial compensation
for participation in this study.

I understand that there is no obligation to answer any question/participate in any aspect of this
project that I consider invasive, offensive or inappropriate.

I understand that all personal data will be kept strictly confidential and that all information will be
coded so that my name and my student's name are not associated with my answers. This consent
form will be kept separate from the questionnaires.

I understand that only the researchers named above and their lab assistants will have access to the
data.

Educator/EA Name (Please print)


Educator/Educational Assistant Signature


Date Best way to be contacted about researcher visits (i.e., phone number and

good time to call, e-mail address, etc.)

Please sign and date this form and return it to your school's administrative assistant, along with the parental
consent form.


100


This study has been reviewed and approved by the Brock Research Ethics Board. (File # 99-342)
If you have any questions or concerns about yours or your child's participation in the study, you may
contact Elizabeth Holmes at (905) 688-5550, ext. 4419, ext. 4519 or Professor Teena Willoughby at (905)
688-5550, ext. 4067. Feedback about the use of the data collected will be available during the month of
February, 2001. A written explanation will be provided for you upon request. Thank you for your help!
Please take one copy of this form with you for further reference .

I have fully explained the procedures of this study to the above volunteer.

Researcher Signature Date


101

Appendix B3

Participant Background Information

1 . What is your child' s age? years months old.

Grade? Sex?

2. To what diagnostic category does your child belong (please check one)?

Autism

Pervasive Developmental Disorder

Rett's Disorder

Childhood Disintegrative Disorder

Asperger's Disorder

Other (please specify)


3. When was your child first diagnosed?


4. Did your child receive any other diagnoses?
Yes

No

If you answered "yes," please specify what they are


5. What is your child's cultural or ethnic background (e.g., French, Chinese, Aboriginal,
African Canadian, etc.)?


6. What is (are) the language(s) often spoken in your home?


7. Please describe the nature of your child's program at school (e.g., is there an
educational assistant present, how many hours he/she attends school, etc.)


)-fi/


102


8. Your child lives with (please check one)

Both parents

One parent

Other (please specify)


9. What is your relationship to this child (e.g., mother, father, etc.)?


10. Does one parent/guardian stay at home with the child? Who?


1 1 . Does one parent/guardian drive the child to school? Who?


12. Does one of you regularly meet with the child's teacher or educational assistant?
Who?


13. What activities do you usually do with your child?


14. Does your child have any siblings?


No


Yes


If you answered "yes", please list how many and their ages.

1) 2)

3) 4)

5) 6)


103


15. What is the highest level of education you have completed (please check one)?
Elementary school Ph.D.


Junior High school post-Doctoral

High school other (please specify)

Community college

University

MA

16. If you have a spouse, what is the highest level of education he/she has completed
(please check one)?

Elementary school Ph.D.


Junior High school post-Doctoral

High school other (please specify)

Community college

University

MA


104
Appendix B4

Preschool Play Behaviour Scale (Coplan & Rubin, 1998)


The following statements concern your opinions about your child's play behaviours when
he/she is in your care (as opposed to when he/she is at school, with another caregiver,
etc.). With respect to your child's play habits, indicate the extent to which each statement
is applicable by circling the number which best expresses your opinion or feeling. Please
try to compare your child to other children of the same age, perhaps peers of his/hers you
have observed at school or in the community. Although it is true that children's
behaviours may be quite variable, please try to make a general evaluation of your child's
"everyday" behaviours.

1 . Talks to other children during play.

1 2 3 4 5

never hardly ever sometimes often very often

2. Plays by himself/herself, examining an object or toy.

1 2 3 4 5

never hardly ever sometimes often very often

3. Plays "rough-and-tumble" with other children.

1 2 3 4 5

never hardly ever sometimes often very often

4. Plays in a repetitive fashion.

1 2 3 4 5

never hardly ever sometimes often very often

5. Plays "make-believe" with other children.

1 2 3 4 5

never hardly ever sometimes often very often

6. Engages in group play.

1 2 3 4 5

never hardly ever sometimes often very often


105


7. Engages in pretend play by himselCherself.

1 2 3 4 5

never hardly ever sometimes often very often

8. Plays alone, building things with blocks and/or other toys.

1 2 3 4 5

never hardly ever sometimes often very often

9. Wanders around aimlessly.

1 2 3 4 5

never hardly ever sometimes often very often

10. Plays in groups with (not just beside) other children.

1 2 3 4 5

never hardly ever sometimes often very often

1 1 . Plays "make-believe," but not with other children.

1 2 3 4 5

never hardly ever sometimes often very often

12. Watches, or hstens to, other children without trying to join in.

1 2 3 4 5

never hardly ever sometimes often very often

13. Prefers junk items to appropriate toys.

1 2 3 4 5

never hardly ever sometimes often very often

14. Plays by himselC'herself, drawing, painting pictures or doing puzzles.

1 2 3 4 5

never hardly ever sometimes often very often


106


15. Engages in active conversations with other children during play.

12 3 4 5

never hardly ever sometimes often very often

16. Prefers to play with the same toy all the time, even when other toys are offered.

1 2 3 4 5

never hardly ever sometimes often very often

17. Plays alone, exploring toys or objects, trying to figure out how they work.

1 2 3 4 5

never hardly ever sometimes often very often

18. Remains alone and unoccupied, perhaps staring off into space.

1 2 3 4 5

never hardly ever sometimes often very often

19. Loses interest in new toys quickly.

1 2 3 4 5

never hardly ever sometimes often very often

20. Engages in pretend play with other children.

1 2 3 4 5

never hardly ever sometimes often very often

21. Engages in playful/mock fighting with other children.

1 2 3 4 5

never hardly ever sometimes often very often

22. Chooses toys based on the sensory stimulation they might provide.

1 2 3 4 5

never hardly ever sometimes often very often


dO!


107


23. Takes on the role of onlooker or spectator.

12 3 4 5

never hardly ever sometimes often very often

24. Plays more appropriately when prompted to do so.

1 2 3 4 5

never hardly ever sometimes often very often


108

Appendix B5

Parent/Teacher Interview Protocol
I am going to ask you a few questions about your child/student's play behaviours.
Please think of your child/student's typical play behaviours, as opposed to behaviours
that occur only occasionally. If you do not understand a question, please do not hesitate to
ask me to rephrase it or provide you with examples of appropriate answers. Remember
that you do not have to answer any question that makes you feel uncomfortable.

1 . Please tell me if your child/student engages in any of the following behaviours:

-plays with an object without taking its functional features into account (e.g., banging a
toy truck on the floor or putting it in his/her mouth instead of running it along the floor)

Yes No Don't know


-plays with an object as its function denotes (e.g., running the toy truck along the floor)

Yes No Don't know

-play that is not dependent on the presence of a physical object (e.g., pretending to drink
using an invisible cup or bottle)

Yes No Don't know


-use of items to represent absent objects (e.g., using a wooden block to represent a bottle)

Yes No Don't know

-pretending an object possesses qualities it does not have (e.g., pretending a carpet is
"magic")


Yes No Don't know


-role-playing (e.g., pretending to be a parent or a teacher)

Yes No Don't know


109


-building structures with toys (e.g., wooden block towers, sand castles, etc.)
Yes No Don't know


-playing games that have rules

Yes No Don't know_

-playing alone, away from others

Yes No Don't know_

-playing alone, but near others

Yes No Don't know

-playing with others with a common purpose to the activity

Yes No Don't know


-staring into space while others are playing

Yes No Don't know_

-wandering around while others are playing

Yes No Don't know_

-watching others play, but not joining them

Yes No Don't know


2. Are there any other types of play behaviours your child/student engages in at
home/school that were not mentioned already?


no


3 a) Does your child have a particular toy that he/she plays with more often than all the
others?

Yes (answer questions 1-4) No (answer question 5)

b) Could you describe it, please?


c) Does this toy provide your child with some sort of sensory stimulation (i.e., is it
pleasant to see, hear, smell, taste, or touch)?


4. What kinds of behaviours does your child engage in with this favourite toy (e.g.,
spinning, repetition, etc.)?


5. Does your child grow upset if this toy should go missing or be taken from him/her?
Yes No

6. When your child is upset about something, does he/she seek out that toy for comfort?
Yes No

7a) Does your child like to play with his/her peers?

Yes No

b) Is there a particular playmate he/she prefers?

Yes No


Ill


c) What is his/her relationship to this person (e.g., sibling, classmate, etc.)?


d) Please describe how they play together.


8a). Does your child/student grow upset easily when asked to play with others?

Yes No

b) If yes, what kinds of things must be done to calm him/her down?


9. Do you feel that your child/student is at, above, or below the academic level of
his/her peers?

At the same level

Below

Above


10. Do you believe that your child has the same opportunities to play as his/her
typically-developing peers? Why or why not?


112


1 1 . What do you think your role is in structuring your child's play time?


Thanks very much for your time.


113


Appendix B6


Coder:


Participant Code:


Behavioural Categories on the Play Observation Scale (Rubin & Coplan, 1998)

Time Sample

1

2

3

4

5

6

7

8

9

10

11

12

Transitional

Unoccupied

Onlooker

Constructive

Exploratory

Functional

Dramatic

Games

Reading

Gross Motor (e.g., running, dancing, etc.)

Small Motor (e.g., cut & paste, colouring, etc.)

Listening Centre

Computer

Left classroom

Snack

Peer Conversation

Adult Interaction

Double Coded Behaviours:

Anxious behaviours

Aggression

Rough-and-tumble

Crying

Socially Inappropriate Behaviour

Laughing

Helping Peer

Peer Re-directs or Motivates Participant

Peer-initiated Play

Participant-initiated Play

Talking to self/toys

Perseveration

Solitary

Parallel

Group


Notes:


114
Appendix CI

BROCK UNIVERSITY, DEPARTMENT OF PSYCHOLOGY

Parental Debriefing Form
"A Naturalistic Observation of the Play Behaviour of Children with Autism"

We would like to express our appreciation for your participation in the present
research. At this point, we would like to provide you with some information about this
study.

The intent and purpose of the study were exactly as they appeared. There was no
concealment of information from yourself or your child. It was our intention to observe
and record the types of play engaged in by children with autism during free play time at
school. The research question this study will attempt to answer is: During free play time
at school, at what cognitive and social levels do children with autism play?

As a parent, you were asked to complete two questionnaires asking for
background information about your child and about how he/she plays when under your
care. This was done in order to collect statistical data as well as to control for differences
between the participants when looking at the results. You were also verbally interviewed
regarding what types of play you believed your child to be capable of This was done to
compare parent and educator perspectives.

If you wish to receive a copy of the results of this study, you can telephone or e-
mail Elizabeth Holmes using the contact information listed below. Results should be
available by early in the year 2001. Alternatively, you may leave your own contact
information with one of the researchers.

Again, thank you for your participation in this study. Any fijrther questions you
might have can be addressed to;

Elizabeth Holmes, MA Candidate, Department of Psychology, Brock University
Ph: 688-5550, ext. 4419, ext. 4519 E-mail: eh99ae@badger.ac.brocku.ca

Prof Teena Willoughby, Ph.D., Department of Psychology and Child and Youth Studies,

Brock University

Ph: 688-5550, ext. 4067 E-mail: twilloug@spartan.ac.brocku.ca


115
Appendix C2

BROCK UNIVERSITY, DEPARTMENT OF PSYCHOLOGY

Educator Debriefing Form
"A Naturalistic Observation of the Play Behaviour of Children with Autism"

We would like to express our appreciation for your participation in the present
research. At this point, we would like to provide you with some information about this
study.

The intent and purpose of the study were exactly as they appeared. There was no
concealment of information from yourself or your student. It was our intention to observe
and record the types of play engaged in by children with autism during free play time at
school. The research question this study will attempt to answer is: During free play time
at school, at what cognitive and social levels do children with autism play?

As a teacher, you were asked to complete a questionnaire asking about how your
student plays when under your care. This was done in order to collect statistical data. You
were also verbally interviewed regarding what types of play you believed your student to
be capable of This was done to compare parent and educator perspectives.

If you wish to receive a copy of the results of this study, you can telephone or e-
mail Elizabeth Holmes using the contact information listed below. Results should be
available by early in the year 2001. Alternatively, you may leave your own contact
information with one of the researchers.

Again, thank you for your participation in this study. Any further questions you
might have can be addressed to:

Elizabeth Holmes, MA Candidate, Department of Psychology, Brock University
Ph: 688-5550, ext. 4419, ext. 4519 E-mail: eh99ae@badger.ac.brocku.ca

Prof Teena Willoughby, Ph.D., Department of Psychology and Child and Youth Studies,

Brock University

Ph: 688-5550, ext. 4067 E-mail: twilloug@spartan.ac.brocku.ca


116


Appendix C3


BROCK UNIVERSITY, DEPARTMENT OF PSYCHOLOGY

Feedback Letter

Title of Study: "A Naturalistic Observation of the Play Behaviour of Children with

Autism"
Researchers: Elizabeth Holmes, Graduate Student, Psychology Department, Brock

University, 688-5550 ext. 4419, ext. 4519

Teena Willoughby, Thesis Supervisor, Psychology Department, Brock

University, 688-5550 ext. 4067

To the parents and educators of the research participants:

As promised, here are the results of the study your child/student participated in. Please do
not hesitate to contact the researchers listed above if anything is not clear to you.

Demographic Variables .

Play information provided by parents .

Play information provided by educators .

Observed Play Behaviours .

Thanks again.
The researchers.


117


Appendix D

Regrouping of PPBS Items


New Variable


PPBS Questions


Peer Conversation

Solitary-exploratory

Rough-and-tumble

Group-dramatic

Group

Solitary-dramatic

Solitary-constructive

Unoccupied

Onlooker
Solitary-fiinctional


1 . Talks to other children during play

15. Engages in active conversation with other children
during play.

2. Plays by himselC^herself, examining an object or toy

17. Plays alone, exploring toys or objects, trying to figure out
how they work

3. Plays "rough-and-tumble" with other children.

21. Engages in playfiil/mock fighting with other children.

5. Plays "make believe" with other children.
20. Engages in pretend play with other children.

6. Engages in group play.

10. Plays in groups with (not just beside) other children.

7. Engages in pretend play by himselfi'herself

1 1 . Plays "make-believe," but not with other children.

8. Plays alone, building things with blocks and/or toys.

9. Wanders around aimlessly.

18. Remains alone and unoccupied, perhaps staring off into
space.

23. Takes on the role of onlooker or spectator.

14. Plays by himself, drawing, painting pictures or doing
puzzles.