Efficacy of Gestalt Visual Motor Integration Learning (GVMIL) to Improve Visual Perception in Children with Autism

2-1 Background

Children with autism generally show atypical cognition, which is sustained or becomes worse unless children receive proper therapy or interventions. Thus, it is important to provide children with proper interventions to develop typical cognition and behaviors. Among many different methods of intervention, art-based intervention is known to be effective in enforcing active cognition [19]. Contending that visual perception is a cognitive activity, Arnheim [20] argued that artistic expression is a method of visual communication. This concept allows room for improving communication with artistic activity. More importantly, art-based intervention can not only reduce atypical behaviors by improving visual perception but can also improve sociality by visual communication.

"Form" in art implies a structure that is systemized by visual perception. In particular, people typically recognize the whole rather than parts. By viewing the essence of an object as a systemized whole, gestalt psychology insists that people tend to perceive something simply and symmetrically, which is considered to be the basic principle of human visual perception. Tacca [21] argued that visual perception and cognition are similar in terms of their structural property of systematicity. Thus, art activities associated with visual perception and gestalt psychology may have the potential to improve visual perception during childhood development. Visual perception is not only reinforced by art intervention but can also be reinforced by imitating body motion. Interventions exploiting imitation have successfully improved the behavior of children with autism as well as young children without ASD [22]. Han [23] argued that an intervention with body motion integrated into sensory training could significantly influence the development of visual perception and the sense of movement in children with developmental disabilities.

Intervention evolves with advances in technology, and new forms of media have the potential to be cost-effective educational materials. In particular, new media can enrich interactivity using digital technology and networks. Whereas conventional multimedia interventions use motion pictures with children, such media are limited in realizing interactivity, which is generally rendered possible by a teacher or parents. Brown [24] contended that digital media enable behavior to induce knowledge so that problems can be solved by integrated thinking. Recently, growing interest in edutainment content exploiting new media to enforce creativity has enhanced the importance of self-motivated learning. Edutainment has the potential to generate self-motivation and enforce interaction through active participation. Exploiting this property, new media intervention is a promising method to improve visual perception with personalization in a cost-effective manner.

2-2 A prototype of edutainment content with motion recognition

The interactivity and natural user interface of edutainment present new possibilities for intervention in children with autism. In prior work [25], we developed a prototype of edutainment content for recognizing body posture in response to a presented image. Because the proposed intervention program is based on that prototype, we reproduced some important results from the prior work. The prototype uses Kinect to recognize the natural motion of children and is designed to allow children to imitate what is recognized with body movement. The prototype also facilitates collaboration among children by completing a given mission together.

The prototype comprises three sections, introduction, main body, and ending, which are played in sequence. The introduction is an animated motion picture that asks participants to follow the motion of animated characters and explains what will occur in the main section. Children are requested to fit their bodies to figures presented on a screen or monitor in the main section. When the child and the figure are matched to a high degree, some other figures are continuously presented several times. In the final section, animated characters compliment what the children have achieved, causing the children to feel accomplished. Characters are also developed to make children feel intimate and build strong bonds with the content. The characters are colored with high color saturation, bright hues and asymmetric body shapes in which the head is larger than the torso. The characters are designed by combining the basic figures of circles, triangles, and rectangles with pandas, foxes, and owls, respectively. These animals have some characteristics in common with each basic geometric figure. Their names are easily pronounced by young children and are associated with the name of the basic figure. In a pilot test conducted with 5-year-olds, the children actively participated and enjoyed this content, which demonstrated its potential to serve as edutainment content.

2-3 GVMIL Program

The prototype of edutainment content developed in the prior work was used to create an intervention program for young children with autism. The program, which is called gestalt visual motor integration learning (GVMIL), is designed to improve visual perception by imitating objects that are the basic elements of formal composition, such as plants, animals and transportation. The point, line, and plane approach can help children with autism understand how the whole can be constructed from parts to form gestalt. In this approach, connecting points becomes a line; a collection of lines becomes a plane; and a collection of points, lines, and planes compose the image of an object. Defining visual thinking as visual perception integrated with thinking, Arnheim [20] argued that visual perception allows human beings to truly understand experience. Experiencing the parts of an object by embodied cognition may help children with autism form the concept of the whole, which may induce improved cognition.

The GVMIL program also uses the inclusion and awareness characteristics of gestalt therapy to cause children with autism to experience and express gestalt. Inclusion is naturally immersing oneself as fully as possible into the experience of the other while simultaneously retaining a sense of one's autonomous presence [26]. Similarly, awareness, or being in touch with one's own existence, exists on a continuum with both ends, foreground and background. Once awareness is fully experienced as gestalt in the foreground, awareness can be integrated into the background. Analyzing eye tracking, Klin et al. [27] contended that weak social interaction is exacerbated by attention to non-salient features in the background rather than a stimulus associated with social cognition. Allowing children to imitate objects in the foreground and background may provide an opportunity to properly improve shifting focus. This program introduces inclusion to allow children with autism to experience foreground and background, which may help these children experience gestalt. Shifting between the foreground and background in gestalt therapy is embodied cognition, which can occur in the entire person [28]. Motion following visual perception in GVMIL provides an opportunity for participants to experience embodied cognition.

GVMIL is also designed to encourage collaboration among participants. Adams and Robinson [22] argued that Kinect enables children to play in teams, which facilitates children cooperating with one another. Bianchi-Berthouze, Kim, and Patel [29] argued that increased body movement in an educational game can increase engagement and social experience through multi-plays. Young children in a small group with GVMIL intervention may have to collaborate to fit their bodies onto the images of the objects presented. As a result, GVMIL can increase natural interaction among children, which can provide an opportunity to improve sociality. GVMIL can also be applied to integrated education, in which students from special education can be integrated into regular classrooms. Children with autism may have a better chance of naturally imitating the behaviors of typically developing children by integrated education.

GVMIL comprises 12 sections; each group of 4 sections reflects a different aspect of gestalt psychology and therapy. The sections are constructed in the order of experiencing the gestalt principles of inclusion and awareness and the formation of gestalt. The first group of 4 sections addresses gestalt principles such as proximity, similarity, continuation, and closure, which are primarily expressed with points and lines to train typical basic visual perception with the basic elements of formal composition. Children learn by matching the outlines of images with their body posture in the middle 4 sections. These sections attempt to realize inclusion in gestalt therapy by encouraging children to imitate the form of an object with their bodily motions. The experience of the human being is determined by the gestalt into which sections of the environment are subjectively structured [6]. GVMIL encourages participants to focus on an outline that is an articulated and simplified whole and is designed to be played collaboratively among participants to guide natural body contact to improve sociality. The visual perception of given outlines can elicit association with corresponding objects. Thus, the images of familiar objects, such as writing materials in a classroom, plants, and animals, are used.

The last 4 sections are designed to help children understand the concepts of foreground and background and to experience gestalt by playing with those concepts. GVMIL forces participants to experience the foreground, which may result in an awareness that engenders a change in cognition. Each section is articulated with a unique theme and multiple outlines of different objects. In section 9, participants are requested to select the image corresponding to the parts of the body in their peers. If the body parts are selected correctly, the outline of the corresponding image is shown on screen to be played. The theme of section 10 is animal houses. The outlines of an animal, such as a cat, fish, or bird, and the house of the animal are presented in different sizes to illustrate the concept of foreground and background. When the animal and the house are successfully matched with a body pose, a short motion picture including animals and their houses is played so that both the animal and house can be recognized in a more realistic context, which may help children with autism to properly identify typically perceived objects. Section 11 is designed with the same structure but with the theme of transportation. In the last section, participants are asked to form an image freely to enforce imagination and collaboration. An instant change of a collaboratively formed image with the children's motion may present an opportunity to experience embodied cognition and gestalt. The goal and content in each section of the GVMIL program are summarized in Table 1. Some screen shots of the GVMIL program are presented in Figure 1 to present more concrete information on this program.

Table 1.

Summary of the GVMIL Program

Fig. 1.

Sample Screen Shots of the GVMIL Program (The first column: Intro Section, The second column: Main body, The third column: Concluding section)

GVMIL also satisfies all four of the principles of therapeutic educational intervention proposed by Cain and Seeman [30]: repetitive exercise; personalization; a combination of visual, auditory, and kinesiology stimuli; and step-wise activities with frequent feedback. GVMIL also includes some desirable properties defined by Parsons & Mitchell [31] for social skills training, such as the repetition of a target skill, ease of use for schools, and affordability for the home environment. From the observation that children with autism may prefer objects producing a sensory effect such as light or sound for imitating actions [12], the sound "oh-Yes" was developed to inspire children to participate when they successfully match their bodies to the image of a presented object.

3-1 Material

Software for GVMIL was developed separately for each section. GVMIL was installed on the desktop computer located in the classroom in which the intervention occurred. The computer was connected to a speaker and projector so that the children could play while watching a large screen rather than a computer monitor. Kinect was also connected to the computer so that it could feed information on the body parts and the joint positions of each participant to recognize the motions of each participant in the software for GVMIL. Because the software included an introduction, a main body, and a concluding section all together, each session began with clicking a starting button and ended when the motion picture in the conclusion was over. No additional materials were used for this intervention.

3-2 Participants

A typically developing 5-year-old boy, a typically developing5-year-old girl, and two 7-year-olds with autism formed a group to apply the intervention in the form of integrated education. This intervention was executed with the consent of the parents of the children participating in the intervention. Because the children with autism had previously been diagnosed to qualify for a government welfare benefit, additional diagnosis was not conducted. Throughout this paper, the two children with autism are referred to as Aron and Billy. Aron often talks with friends without looking at their faces although he makes eye contact for a short time if necessary. Aron also tends not to listen to peers and takes comments out of context. Aron can also easily express his negative opinion about what he does not like and has a short attention span, even for preferred activities. Aron often wanders around the classroom by himself rather than playing with other children. Billy appears to have milder symptoms than Aron. When calling to his friend, Billy makes eye contact for a short period and then turns his head away. Billy knows the names and faces of every classmate and can understand statements such as "Stop it" and "Don't do it", although Billy tends to continue the behavior and generally does not follow orders and rules. Billy can focus on preferred activities for a long time and tends to avoid non-preferred activities.

3-3 Analysis

The primary purpose of this intervention, i.e., improvement in visual perception, was quantitatively analyzed. We also made some qualitative observations on visual perception and sociality, which are implicitly expected from the relation between cognition and sociality. The motor-free visual perception test-3 (MVPT-3) [32] was selected to evaluate the degree of visual perception. Although there are other tests for visual perception, such as the Bender visual motor gestalt (BGT) [33] and the developmental test of visual-motor integration 6 (VMI-6) [34], those tests include motor integration to test visual perception, and their accuracy can largely depend on the skills of the examiner and the degree to which the children understand the instructions. Conversely, the MVPT-3 does not require motor integration because this test comprises multiple choice-type questions. The MVPT-3 is likely to be the most appropriate test for the evaluation of this intervention because we are interested in visual perception rather than motor skills.

4-1 Quantitative Analysis of Visual Perception

The results of evaluating visual perception with the MVPT-3 in the baseline phase are summarized in Table 2. When comparing the effects of the intervention on visual perception between children with autism and typically developing children, the test results of typically developing children named Carol and David were included. The age equivalences of the visual perception ability of children with autism are substantially lower than their chronological ages, whereas the age equivalences for typically developing children are marginally higher than their chronological ages. Table 3 summarizes the MVPT-3 results obtained post-intervention. All four children improved their raw scores on the MVPT-3, although Aron's age equivalence remained less than 4 years old, which is the minimum equivalent age measurable with the MVPT-3. More significant improvement in raw scores for children with autism may suggest that the GVMIL program can be more effective for children with visual perception issues. Billy demonstrated the greatest improvement in age equivalence, i.e., 1 year and 8 months over less than 4 chronological months, and David showed similar improvement. These results support the concept that GVMIL can be a promising intervention for improving the visual perception of children with autism and possibly even for children without autism.

Table 2.

MVPT-3 Results for Baseline Phase

Table 3.

MVPT-3 Results for Post-intervention

4-2 Qualitative Observations

Aron did not show much interest in the GVMIL intervention at the beginning. Often, Aron stood without trying to play with the GVMIL or played reluctantly with other children, without much motivation. However, Aron was observed to make some progress in participating more spontaneously after session 8. After session 15, Aron began to show interest in the intervention session even during the normal daily activity session and was also observed to make natural physical contact with other children and even ask for the opinions of other children regarding some issues while actively participating in the intervention sessions.

Billy was responsive to new stimuli in the intervention program, although he did not actively participate in the intervention. Billy enjoyed making the image of himself appear on the screen in the beginning. After becoming accustomed to new stimuli, Billy appeared to lose interest in this intervention. He generally focused for a short period of time with the help of other participating children while often wandering around the classroom during the intervention session. Although Billy was not interested in this intervention for the entire session, he did exhibit interest in the activities of other children during the intervention. Billy also appeared to be more comfortable with physical contact with other children as the intervention continued.

Many different types of interactions occurred between the typically developing children and the children with autism. In the beginning, the typically developing children often attempted to partner with Aron because interacting with Billy was more difficult. However, as the intervention continued, the children tended to attempt to help whichever friend was sitting next to them without discriminating. Although the children were awkward playing together in the beginning, the typically developing children began to help Aron and Billy become properly positioned to fill in the outlines in the intervention. Whereas one typically developing child showed a tendency to monopolize activities during the intervention, the other child was actively attempting to help Aron and Billy. In addition, Aron and Billy were observed to imitate the behaviors of Carol and David and share emotions with the typically developing children.

5-1 Visual Perception

Pellicano [35] showed through longitudinal research that weak central coherence associated with local processing did not significantly improve over the course of 3 years, whereas capacities for empathy and executive functions among children with autism showed significant changes. The GVMIL program, which led to improvements in visual perception, may represent a promising method in this respect, although the program requires additional evidence and longitudinal research. Although body posture in the GVMIL program is used as a method of realizing embodied cognition to improve visual perception, one may consider body posture to be an intervention for visual-motor integration. Ayres [36] developed sensory integration therapy to enforce the relation between sensory experience and motor activities. However, in our research, visual motor integration was not the goal, but rather the means. Moreover, Dawson and Watling [37] argued that sensory motor therapies have not been properly validated despite numerous associated interventions. Although we used the method of integrating vision with body posture, we do not know which sensory integration is the most useful for improving cognition; this question requires further attention.

Significant improvement in visual perception with the proposed intervention may serve as evidence for the importance of interventions to support proper cognition at a young age. Pellicano [35] argued that bias in local processing emerges early during the course of development. Despite the efficacy of the proposed intervention, more refined articulation adapted to the level of development in cognition and age may render this type of intervention more effective.

5-2 Social Skills

There is a close association between sociality and cognition. Kaiser and Shiffrar [38] suggested two different views explaining how visual perception and social cognition in autism are related. People with autism may have typical visual perception and may process perceived information inadequately. Conversely, people with autism may have abnormal visual perception, whereas their social cognition processing works well. From the perspective of cognition as embedded in the experience resulting from the effect of body motion on surrounding environment, social cognition can be considered experience associated with actions connected with social interaction [27]. Information produced by cognition has significant effects on human behavior, whereas sociality forms behavior and cognition [39]. Sociality builds on the capability of social perception, such as recognizing subtle emotional cues and picking up relevant information.

Bauminger [40] defined social cognition as the ability to read spontaneously and interpret social and emotional cues correctly. In terms of this definition, some previously unseen social behaviors resulting from social cognition were noted. Moreover, children appeared to share emotions with peers in some cases. These findings highlight that the GVMIL program may help to improve sociality to a degree. Intervention to improve perception has been reported to be more important than intervention directly remediating social behaviors [5]. The GVMIL program does not teach specific skills; rather, the program attempts to strengthen the foundation for proper behavior by enforcing typical visual perception.

5-3 Educational Implications

One of the essential implications of this research is that visual perception in autism can be improved by natural play. In addition to emotional development, cognitive development is required for social play. Cognitive development is known to be associated with moving through the sensory exploration of objects [41]. The GVMIL fosters collaborative play by having children match presented outlines with combinations of their own body posture. From this perspective, the GVMIL may enforce social play during the intervention. Children with autism may learn new skills from peers and share the joy of achievement from collaboration, which is implicitly provided by the GVMIL.

Improved visual perception may imply a reduction in a bias toward local processing, which results in cognitive functions. The GVMIL may help to change an unusual cognitive style to allow children with autism to integrate information in a more relevant context. Milne and Griffiths [42] noted that children with autism demonstrate a literal understanding of spoken language without being overly influenced by sentence context. Improvement in cognition through GVMIL intervention may have a latent effect on language processing in autism, and improving visual perception is expected to have consequential effects on the development of important behaviors required for typical interaction. The pre-requisite for proper social behavior is likely to be having proper information. Indeed, many human behaviors are based on visual information. In this regard, GVMIL intervention may influence social behaviors such as reading social cues, expression of emotion and establishing shared attention. The GVMIL may also contribute to learning complex motor skills and having appropriate body posture.

Although we could not determine which function in visual perception was further developed with this intervention, interpreting improvement in visual perception as being a result of having more proper information has several educational implications in social behaviors and learning. Maynard [43] contended that although social interactions involve concerted actions having an intrinsic gestalt structure, people with autism often have more locally organized orientation and sensibility driven by the local analysis of social information. Joint attention is considered to be important in the sense of a shared experience [41], and joint attention as a cognitive process is strongly associated with available information. This connection implies that developing visual perception may improve joint attention and help to develop a sense of shared experience. Vetrayan, Zin, and Paulraj [44] showed the strong relation between imitation and visual perception using motor-reduced visual perception, the unstructured gesture and object imitation in the school functions of children with autism. From this result, the authors argued that visual perception substantially affects the skills required to function in school such as paying attention and reading. This result implies that the GVMIL may improve school functions, which can enhance the efficacy of other education and develop the new learning behavior.

5-4 The GVMIL Program

Pares et al. [45] contended that children with autism cannot be typified because of the wide spectrum of disorders covered by autism, although users must be typified in the design of an interactive application. These authors designed an interactive space called MEDIATE with the primary goal of creating a playground in which children with autism could have fun. They only used abstract images to remove the possible dependency on the content represented in an image rather than the space itself. Similarly, the target users of the GVMIL program were young children with autism. However, we used both abstract images and representational images. With the help of friendly-looking characters, children showed only positive responses to the presented images. However, because images can occasionally be problematic, images causing a negative response must be replaced by consultation before beginning an intervention.

Infants are known to develop the basic communication skills required for proper social interactions during the first two years of life [46], and the current GVMIL program may not be appropriate for infants. Thus, refined adaptation for infants is needed to render the program applicable for more efficient intervention in younger subjects. Applying Kansei engineering to improve the emotional quality of GVMIL can be a candidate approach to make it realized efficiently as it has successfully improved interaction of mobile web [47]. .