Abstract

Interactive learning tools developed at the Center for Spoken Language Research at the University of Colorado have been applied effectively to language training to help profoundly deaf children learn and use new vocabulary in their daily lives. The language training tools, which incorporate speech recognition, speech generation and facial animation technologies in multimedia applications, are easy to learn to use, easy to use, and have great potential to benefit children with cognitive disabilities. We are conducting a pilot study this summer to develop and test applications to benefit children with ASD—Autism Spectrum Disorders. The pilot study brings together researchers and software developers, autism experts, teachers, parents and children with ASD. The team is building and testing a variety of applications designed to meet the specific needs of children with ASD. These applications will be ported to classrooms and resource rooms in the special education program during the 2001-2002 school year, and to children’s homes for additional learning outside of the school environment.

Rationale for Pilot Study

Autism Spectrum Disorders (ASD) encompass a wide spectrum of cognitive and behavioral profiles, which include the diagnoses of autism, Asperger syndrome and pervasive developmental disorder-not otherwise, specified (PDD-NOS). Children on this spectrum are quite heterogeneous with cognitive abilities that can range from mental retardation to near genius IQ.

Common characteristics for children with ASD include:

• Difficulty with social interaction
• Need for predictability, consistency and routine
• Difficulty reading other’s emotions and intentions
• Difficulty generalizing skills from one activity to another
• Difficulty with transitions

The children with autism also have significant language delays and differences.

Because of cognitive, memory and attention constraints, many children on the autism spectrum need to “over learn” information for it to be automatic and permanent. Teachers and paraeducators rarely, if ever, have the time to teach at the intensity that these children need. The current vocabulary tutor and additions to the tutor could fulfill the need for intense and repetitive instruction. The animated character can continually teach, never act bored, review if the child forgets information, wait patiently as the child processes information, repeat for clarification and give consistent reinforcement.

Also, because these children have substantial difficulty generalizing from one situation to another, children on the autism spectrum need to learn from pictures and situations that are relevant to their environment. Too many programs are canned and teach from pictures and situations that the students cannot generalize from (i.e. they teach about common household items but not the ones that the child is familiar with). The vocabulary tutor can teach from pictures that are specific to the child’s environment.

Most of the children on the autism spectrum spend part if not most of their day in general education classes. There are certain routines, expectations and topics that all children are expected to know. The child with ASD often needs explicit instructions on these routines that other children have learned implicitly. Also, many children with ASD need to be pretaught about a specific topic before the classroom teacher presents it to the rest of the class. This allows the student with ASD to have more cognitive resources available for learning conceptual information instead of expending resources wondering about context and vocabulary. With minimal effort and minor adaptations of existing software, tutors can be set up to teach about expectations, routines and to preteach academic information.

Proposed Work

CSLR has assembled a team that is working together this summer (2001) to explore and pilot tutors for young children with autism and Asperger. The team includes Scott Schwartz, Ph.D. who teaches Autism Spectrum Disorders through the Speech Language and Hearing Sciences Department at C.U. Scott is the principal investigator on this pilot project. He is also an autism expert and a special education consultant for the Boulder Valley School District.

Emilie White, who has worked in the field of autism for many years, is assessing needs, creating programs that work best for children on the autism spectrum and implementing those programs. Emilie is trained as a behavioral therapist and has also received training in TEACCH. Emilie is also the Family Specialist in the Boulder Valley School District where she provides support and training to families with preschool age children on the autism spectrum.

Linda Corson, Director of Corpus Development and Toolkit Training at CSLR, is working with the team as technical support and software trainer. She has conducted trainings on the vocabulary portion of the toolkit and outreach training will be conducted for parents and schools as they begin using the software at home. Linda is also overseeing the ongoing tutor development and developing data collection tools for the project. Linda has also taken on the responsibility of working with one of our students on higher level computer skills such as programming and creating applications using the Speech Toolkit.

Kim Thompson, a graduate student in Education and Linguistics, is working as a summer intern, developing applications, collecting data, videoing and working directly with the students to implement the programs.

The program looks very different for each child. One child is working on mouse control; others on basic vocabulary in areas relevant to their daily experiences and academic curriculum; others on interpreting and categorizing facial expressions in relation to specific situations, and another student is creating applications and learning the inner workings of the Toolkit.

Goals for the summer project include:

• Determining minimal level of computer skills necessary for accessing computerized tutors;
• Determining best candidates for learning from computerized tutors based on the cognitive and behavioral profiles of the child;
• Determining appropriate content and sequence of tutors for individual children with ASD;
• Determining how to interface tutors with other therapies or programs that the child might be involved in;
• Collecting pilot data on the effectiveness of the tutors for learning;
• Determining strengths and weaknesses of current software for this population;
• Proposing additions to enhance current software so that it is more effective and appropriate for children with ASD; and
• Assessing parents’ needs and their suggestions on how to make the tutor more effective with their child.

Using the information that we gather from the summer project, the following is proposed for the upcoming school year (some of which will occur prior to the beginning of school):

• Improve and enhance current software to increase teaching effectiveness;
• Train teachers and other staff members in the BVSD to author appropriate tutors for their children with ASD;
• Provide ongoing training for teachers and parents to facilitate appropriate tutor development;
• Involve parents in the design and authoring of specific tutors for their child;
• Place computers and tutors in BVSD classrooms specifically for children with an ASD;
• Provide software to parents for carryover of skills and learning at home;
• Exploring the effectiveness of the tutors for use with other disability groups including Down syndrome and other children with cognitive differences;
• Establish a Web site in collaboration with the Colorado HEAT center to distribute applications, and to enable teachers and parents to contribute applications; and
• Involve other school districts in the metro by:
• Training specific teachers on how to use the toolkit software
• Providing hardware and the toolkit software in their classrooms
• Setting up systems to collect, analyze and interpret data
• Providing ongoing support to these teachers on appropriate tutor development
• Having periodic meetings between all teachers and districts involved to share information, report on already developed tutors, brainstorm new ideas, troubleshoot, report successes and challenges, make wish lists and recommendations, and compare notes.

Impact

The pilot study is expected to have substantial impact on the way in which teachers, parents and students work with children with ASD in the Boulder Valley School District. Our goal is to demonstrate that the applications we develop provide critical knowledge and skills to children with ASD; that children enjoy using the applications and applying their newfound knowledge and skills; that teachers and parents will view the tools as an invaluable resource; and that teachers and parents will work with us in the future to improve the software, develop new applications and share their positive experiences with other educators.