Introduction: the Value of distance learning for autistic Treatment
Recent psychological studies have revealed that autistic children can neither reason properly about mental states of themselves and others, nor understand emotions (Leslie, 1987; Perner 1991; Pilowsky, Yirmiya, Arbelle, & Mozes 2000). Autism is a multifactor disorder that is characterized by impaired social interaction and communication, combined with repetitive and stereotyped patterns of behavior, and affects up to 1% of school-aged children in some geographic areas. In this article we are concerned with the strategies of rehabilitation of reasoning to improve communication skills of children with autism.
It has been confirmed by multiple clinical studies that the properly timed treatment is essential for the autistic patient to increase the chance for recovery. An early behavioral intervention is highly beneficial for autistic children. There is an opinion with increasing support by multiple researchers that intensive behavioral intervention (that can be stimulated in distance learning) may result in a dramatic improvement of autistic reasoning (McEachin, Smith, & Lovaas, 1993). From the viewpoint of autism experts who believe that there is no alternative to behavioral intervention (thought of as the only way to facilitate compensatory learning; see, e.g., Frith, 2001; Howlin, 1998), distance learning may be a useful aid for the education of parents and rehabilitation personnel.
Multiple technologies have been suggested to support autistic learning, including playing LEGO (Resn-ick, 1987), video-clips together with a set of dolls, and autonomous mobile robots.Also, asking questions about mental states of the scene characters, textual scenarios, and characters of the works of literature (Galitsky, 2003) is a good assistance to parents and rehabilitation personnel in the proper diagnosis and training of the corrupted autistic reasoning.
Virtual educational facilities introduced in this article are delivered primarily online, that is, by networks such as the Internet or intranets, using asynchronous technologies such as computer conferencing or Web-based technologies especially customized for education (Hietz, 1995; Nasseh, 1999). These learning technologies enable and support active, collaborative learning approaches, and the software provides tools for: designing courses and developing an interactive course syllabus, conducting individual and group learning activities, and recording and accessing evaluations and grades. In cases of autistic distance learning, collaborative learning is essential for communication skills. Using this technology the learners are encouraged to perform introspection to communicate the fact that they are capable of understanding this very capability of communication (Meyen, Lian & Tangen, 1997).
We outline the reasons of the high rehabilitation value of distance learning:
• Autistic children frequently prefer computer-based interaction than inter-human interaction; also, they enjoy unique content available online to impress their peers.
• Exchange of successful treatment strategies between rehabilitation centers is easier.
• It provides access to treatment for remote communities currently lacking rehabilitation professionals in the proximity of their residences.
• The distance learning infrastructure, augmented with inter-communication framework, would support the interaction between the members of community of autistic children and their parents. This feature on its own is a useful support for developing communication skills of autistic patients, even though it is not the most natural mode of communication.
However, in addition to the requirement of being delivered online, distance learning software for autistic rehabilitation must impress its users by a number of intelligent features. This requirement is due to the fact that in contrast to conventional learning process, such software has to be capable of substituting interaction with humans to a certain degree. Frequently, autistic patients prefer to deal with software agents rather than with humans. These software agents need to demonstrate the reasoning skills, which are expected to be developed by the learners, rather than just to introduce a domain for reasoning.
teaching autistic children to understand the mental world
Building an adequate formal model of the mental world including emotions is important for teaching individuals whose understanding of mental worlds are (genetically) corrupted and need thorough specification. A non-humanized resource, readily acceptable by autistic children, is used to introduce them to the mental world (of humans) via formalized reasoning. The paradox of our methodological principle is that the reasoning in the mental world, usually supposed to be irrational and displayed as an emotion, can nevertheless be considered from an abstract perspective, formalized and used as a training means. This hypothesis (Galitsky, 2002) is used to form the backbone of our rehabilitation strategy to develop the rational and emotional behavior in the real mental world.
Our model of the corruption of human reasoning is based on the supposition that there are a number of standard axioms for mental entities; these axioms are genetically set for normal children and are corrupted in the autistic brain (Galitsky, 2002). The patterns of corruption vary from patient to patient and are correlated with the specifically outlined groups of autistic children. The simplest of such axiom is not see, therefore not know, highlighted in the so-called Sally-Anne test (Baron-Cohen, Leslie, & Frith, 1985).
The children with autism have to acquire these axioms explicitly, by means of direct training, using a variety of ways to introduce intentions, beliefs, knowledge, and other attitudes of agents and relations between these attitudes. Teaching of such axioms may occur in both formal and natural languages; some of the patients prefer the former over the latter. Frequently, autism is not accompanied by learning disabilities, so the patents willingly participate in training programs of distance learning.
Our practical experience shows that using software-based training allows us to hold the attention of autistic patients for much longer periods than the traditional means of one-to-one treatment by a human. Conventional rehabilitation is always associated with the problem of keeping a patient in a responsive mode. Since for the autistic patients the strict rule-based learning is usually much easier than the direct introduction of the various forms of reasoning about the mental world, the latter is achieved via the former. The methodology of using artificial mental reasoning systems for rehabilitation is based on teaching autistic children the “mechanical” forms of reasoning about the mental world, because the attempts to directly introduce the emotional interaction with the others in a natural manner (e.g., teaching by examples, imitating) frequently fail.
exercises for distance learning
In this section we present a set of exercises that are available online and used in rehabilitation centers in cities across Russia (see Figures 1-4). The common goal of these exercises is to encourage a trainee to focus on such mental entities as intention, knowledge, and belief, as well as cheating and pretending, offending and forgiving, and so forth. In accordance to our underlying formal model of the mental world (Galitsky, 2003), an arbitrary mental state or action can be defined in the basis of intention, knowledge, and belief. Autistic children are encouraged to learn these definitions in a formal or informal way, using formulas, user interface controls of the software systems, or in the natural language. Then the software means are used to evaluate how the definitions were acquired, to perform generalization and abstraction of the perceived entities, and to stimulate its application to the real mental world.
Having been explained the entities of offending and forgiving, a patient is encouraged to build his or her definitions of the above basis, using list box controls to choose from the available set of mental actions, temporal parameters, negations, and other modalities.
The mental attitudes of characters are schematically drawn to illustrate their inter-dependence and attachment to an individual (Figure 2). An automated question answering system (Galitsky, 2003) is used to encourage autistic children to focus on mental entities in their questions and answers.
Figure 1. An interactive form illustrating how the entities of offending and forgiving can be expressed using the basic mental attitudes want, know and believe
Figure 2. A scene with characters used to discuss mental attitudes with autistic patients
A learner specifies the initial mental states of agents in either natural or formal language using controls (Figure 3, three columns for respective agents).
Then the automated reasoning system generates the plausible mental actions and computes the resultant mental states (on the bottom). After that, the learner verifies that the generated mental states match his or her expectations.
In the hypotheses formation setting (Figure 4), the subject is notified that an unpleasant taste is added to some cookies in accordance to some rule that is not disclosed. The subject is required to eat all cookies with good (expected) taste and state that the rest of cookies are altered. For the purpose of verification, a subject is encouraged to formulate the revealed rule when done with the cookies.
conclusion
It is well known that various kinds of reasoning capabilities are built up at different speeds for the same patient. Our advice to the parents of autistic children considering participation in a distance learning environment is that training of each component of mental reasoning should be conducted starting from the earlier ages. This is due to the fact that for each mental task there is the age when this task becomes adequate to the current understanding of the mental world by a learner. Therefore, we recommend that autistic patients become involved in distance learning of reasoning, starting from the age when they are able to understand a textual discourse, and keep exercising until a complete (if possible) mental recovery in terms of interacting with other people occurs.
At the next level of our study of the plausibility of the suggested approach, we have explored how acquiring reasoning skills has led autistic learners to a better understanding of others and of themselves. This is the following step after the completion of a distance learning program. The parents need to be aware that it takes much longer and many attempts and failures to apply the formal reasoning skills to the real world.
Figure 3. A user interface for autistic training based on reasoning about mental states
Figure 4. A hungry subject is suggested to eat cookies from the ten plates
However, today we have evidence of a number of patients who have completed the set of exercises soon after they were diagnosed with autism and achieved almost complete recovery three to five years after that. It occurred partially due to the reasoning skills acquired using the reasoning rehabilitation program. Having started evaluating the reasoning-based rehabilitation strategies from 1997, we observed a dramatic improvement of communication skills and overall behavior of more than 75% of patients who successfully completed the rehabilitation program. As a distance learning environment, the program was established in 2003, after preliminary evaluation of the results.
The phenomenon of computational autism has been introduced to explore the theoretical and practical issues of the mental disorders (Computational Autism, n.d.). Our studies show that the abstract formalism of reasoning about mental states, intended to model the human behavior by a technical system, found an emergent application in the simulation of human agents. The reader should consult Graham-Rowe (2002) for a popular analysis of how studying logical constructions assists in the mental development of autistic patients.
We also mention a set of complimentary software systems for autistic rehabilitation (including Baron-Cohen, 2004; Pritikin, 2003) which target the development of social skills directly via the ability to recognize expressions and emotions in others (Jarr-old, 2004). Also, there is a series of software tools for developing associated skills, in particular Music Box (AutismCoach, n.d.). Most of the software systems available at this Web site have a common educational value and do not have the audience of autistic children as a focus. Also, as of2003, no software packages that could impact the learners with intelligent features existed at the repository. Therefore, we believe that the suggested set of exercises accompanied by automated reasoning and natural language understanding components is a unique, valuable, and efficient resource for distance learning.
Finally, it is worth mentioning that not only learners diagnosed with one or another form of autism are recommended to be involved in the distance learning of mental reasoning. A wide spectrum of mentally healthy children may improve their communication skills by completing the presented exercises at an earlier age in comparison to autistic children, starting at five years old. Moreover, such training would advance the mathematical and analytical skills of the learners; usually it is much harder and takes more effort to advance these sorts of skills by more mature students.
key terms
Autism: A developmental disability significantly affecting verbal and nonverbal communications and social interaction, generally evident before age three, that adversely affects a child’s educational performance. Other characteristics often associated with autism are: engagement in repetitive activities and stereotyped movements, resistance to environmental change or change in daily routines, and unusual responses to sensory experiences.
E-Learning: Education via the Internet, network, or standalone computer. Network-enabled transfer of skills and knowledge. E-learning refers to using electronic applications and processes to learn. E-learning applications and processes include Web-based learning, computer-based learning, virtual classrooms, and digital collaboration. Content is delivered via the Internet, intranet/extranet, audio or videotape, satellite TV, and CD-ROM.
Learning Management System (LMS): A software that automates the administration of training events. The LMS registers users, tracks courses in a catalogue, and records data from learners; it also provides reports to management. The database capabilities of the LMS extend to additional functions such as company management, online assessments, personalization, and other resources. Learning management systems administer and track both online and classroom-based learning events, as well as other training processes (these would need to be manually entered into the system for tracking purposes).
Learning Object (LOS): Also called reusable learning objects; not really a set technology, but rather a philosophy for how content can be created and deployed. Learning objects refer to self-contained chunks of training content that can be assembled with other learning objects to create courses and curricula, much the same way a child’s Lego blocks are assembled to create all types of structures. Learning objects are designed to be used in multiple training contexts, aim to increase the flexibility of training, and make updating courses much easier to manage.
Mental State: A mental condition in which the qualities of a state are relatively constant, even though the state itself may be dynamic; “a manic state.”
Self-Paced Training: Training taken at a time and pace determined by the user. Used historically for text or audio/video self-study courses, the term is used by some organizations now to include computer-based, Web-based, and multimedia training.
Synchronous E-Learning: Communication occurs at the same time between individuals, and information is accessed instantly. Examples of synchronous e-learning include real-time chat and video/audio conferencing. Synchronous e-learning can provide instant feedback on a student’s performance and allows the training to be adjusted immediately if needed. The disadvantages of synchronous e-learning are that the training is not self-paced, and the logistics of scheduling, time zones, and student availability need to be managed.
