Using pivotal response training to
facilitate appropriate play in children
with autistic spectrum disorders
Aubyn C. Stahmer
Children’s Hospital and Health Center, San Diego, California
Introduction
Research examining symbolic play in children with autism indicates that
these children exhibit severe deficits in play skills and play complexity (e.g.
Wing et al., 1977; Jarrold et al., 1993). These children are deficient in
appropriate use of toys and have specific impairments in symbolic and
sociodramatic types of play which involve complex symbol use and social
interaction (e.g. Baron-Cohen, 1987; Stahmer, 1995). Children with autism
rarely engage in the types of creative, spontaneous play activities exhibit-
ed in typically developing children. They may engage in repetitive actions
with toys or become obsessive with them or use toys in stereotyped ways
such as acting out a scene from a favourite movie. Because play skills are
important to social integration, and have been correlated with language,
social skills and general cognitive functioning (e.g. Bates et al., 1979;
Sigman and Ungerer, 1984), it is important that children with autism, like
typically developing children, understand and participate in play.
Although there is overwhelming evidence that children with autism are
deficient in their spontaneous use of complex play skills, there is evidence
that children with autism may engage in higher levels of play (such as sym-
bolic play) when it is elicited either verbally or through modelling (Lewis
and Boucher, 1988; McDonough et al., 1996). Autistic children’s lack of
spontaneous play in free play settings is not, therefore, the result of a com-
plete inability to play, but may be due at least in part to the fact that chil-
dren with autism find play difficult, so they experience repeated task
failure. This in turn would lead to frustration and the pervasive lack of
motivation to play which we see in these children (e.g. Schreibman, 1988;
compare also Churchill, 1971; Koegel and Egel, 1979).
Address for correspondence: A. Stahmer, Children’s Toddler School, 3020 Children’s Way MC5042,
San Diego, CA 92123, USA.
© Arnold 1999 0265–6590(99)CT170OA

Research suggests that increasing motivation has a dramatic effect on
children’s learning (e.g. Koegel and Mentis, 1985), and a training tech-
nique called pivotal response training (PRT) has been developed to
increase motivation in children with autism (Koegel et al., 1987). In PRT
the child is allowed to choose activities, they are provided with multiple
exemplars of the behaviour to be acquired within their chosen activity, and
they are systematically rewarded using natural reinforcers. Easy-to-achieve
maintenance tasks are interspersed with tasks involving new learning in
order to maintain the child’s experience of success. PRT has been widely
used to increase the motivation to acquire language skills in children
with autism (Koegel et al., 1987). More recently the technique has proved
to be extremely successful for increasing manipulative play (Stahmer
and Gist, 1997; Stahmer et al., 1998), symbolic play (Stahmer, 1995), and
sociodramatic play (Thorp et al., 1995) in children with autism. Other
behavioural programmes used for teaching play skills to children with
autism have tended to use a limited number of specific scripts and toys.
This has made generalization and maintenance of new skills difficult. This
technique is structured to assist children with understanding expectations
and increasing skill acquisition. The use of multiple examples and
natural reinforcers in PRT is designed to facilitate skill generalization
and maintenance.
Using PRT to teach play skills
PRT has provided interventionists with a naturalistic training method
that has proven structured enough to help the children learn complex
play skills, while still being flexible enough to allow the children to remain
creative in their play.
The following describes the PRT procedures used to facilitate play.
1) Clear instructions and questions. The play task is made clear to the child.
For example, the therapist might ask the child what could be done with
a particular toy. Modelling and the use of maintenance tasks (tasks the
child has already mastered) are used to clarify new tasks early in train-
ing. Care is taken to ensure that the child is attending to the therapist
and the training materials (e.g. Schreibman and Koegel, 1981).
2) Interspersal of maintenance tasks. Maintenance tasks which require
children to play at levels within their competence are interspersed
with training on new tasks which require the child to play at more
30 Child Language Teaching and Therapy