Qualitative outcomes and impact of a robotic intervention on children with autism spectrum disorder: A multiple embedded case study

Abstract

Most studies of social robot interventions for children with autism spectrum disorder have been laboratory experiments focusing on component skills. There is insufficient evidence documenting the qualitative impact of such programmes on social development and participation of children with autism spectrum disorder. This study aimed to identify the qualitative outcomes of a robot-mediated social skills training programme for children with autism spectrum disorder, examine the impact of such programmes on children’s social participation and identify the essential elements of robotic interventions that are conducive to children’s social development. A case study approach with a multiple case study design was adopted. Sixteen children with autism spectrum disorder, aged 5–11 years, were included. Participants received 12 weekly sessions of robot-mediated social skills training. The successful outcomes relating to social participation were identified as enhanced verbal expression, social awareness and emotional reciprocity. The impacts of the programme on personal development were identified as enhanced self-esteem, self-confidence and emotional expression. Robot friendship, the role of the robot as a facilitator and the presence of a human instructor capable of leading the programme were identified as essential elements of the positive changes. The encounter with a social robot was regarded as meaningful and important to the children with autism spectrum disorder.

Keywords

Social robotics autism social development Canadian Occupational Performance Measure participation

Background

Children with autism spectrum disorder (ASD) frequently demonstrate limitations in their ability to interact and communicate with others. Their limitations in social skills may negatively impact essential domains such as academic development, interpersonal relationships, behaviour, mental health and even adult life outcomes. The use of robotic intervention programmes to enhance the social engagement of children with ASD has been developing rapidly. Evidence shows robotic intervention programmes are effective in improving the social skills such as eye contact, verbal initiations and social imitations among children with ASD (Chung, 2020; Shamsuddin et al., 2012). The basis of the robotic intervention programme is built on the positive effects of human–robot interaction (HRI) for children with ASD. HRI is defined as the dynamically changing relationship between humans and assistive robots (Shamsuddin et al., 2012). A humanoid robot is a robot with a human-like body shape to resemble human beings and mostly with a closed-loop robotic system. Assistive interaction can be induced through interactions between humans and humanoid robots. HRI is assistive to children with ASD because the simple, inanimate and anticipatable characteristics of robots are particularly attractive to these children and it seems to be easier for them to handle the interaction with such robots rather than human beings (Shamsuddin et al., 2012). Assistive robots function as a social mediator in triadic interactions and stimulate and provoke interactive and social responses from children with ASD to the robot as well as the human interaction partners (Simut et al., 2016).

Intervention programmes using HRI have been shown to enhance the social participation of children with ASD (Giannopulu & Pradel, 2010; Srinivasan et al., 2013). Studies have shown that robotic intervention programmes improved joint attention between the instructors and children with ASD (Costa et al., 2015). Stereotyped behaviours were found to be suppressed during child–robot interaction (Shamsuddin et al., 2012). The social skills of the children, such as verbal initiation and joint attention, were also observed to improve over a period of time (Robins et al., 2004). The cultivation of basic social skills is important for children with ASD. Furthermore, such skills should be transferable to real-life contexts and ultimately improve the children’s social participation.

From a community-based perspective, robotics, as a type of assistive technology, is an important part of development strategy to enhance participation. Social participation refers to the gaining of social roles through participation in social activities (Dalemans et al., 2008). An encounter with a robot in a training session stimulates excitement and interest in children with ASD which also elicits the spontaneous initiation of social behaviour towards the robot and then to the human interaction partner (Diehl et al., 2012). The simple and inanimate nature of robots creates a sense of predictability and control for children with ASD during the child–robot dyadic interaction (Huijnen et al., 2019). An encounter with a robot, as perceived by children with ASD, can be true and honest and can lead to enabling them to make positive changes and engage meaningfully with other people (Chung, 2018). There is no published evidence documenting the qualitative impact of robotic intervention programmes on the social participation and development of children with ASD. The qualitative outcomes of robotic intervention programmes should be defined and described so that a thorough understanding can be reached regarding the programmes’ effectiveness in aiding the social development of children with ASD.

The present study

This study aimed to explore the process and qualitative outcomes of robotic intervention programmes for children with ASD. Studies have tended to focus on the quantitative outcomes of robotic intervention programmes and have neglected the process and qualitative outcomes. However, the experiences of children with ASD in their respective contexts should not be studied only quantitatively, as these experiences are mostly subjective phenomena. Moreover, both robotic intervention and the social development of children with ASD are multi-faceted constructs. A proper investigation of these constructs requires an in-depth and contextual examination of the children’s experiences and changes. All of the processes, stakeholders and personal and environmental factors have to be considered to elucidate the relationship between the various constructs. Qualitative research is necessary, as it examines life experiences to facilitate understanding of their meaning in context (Fazli et al., 2018). Previous work of the research team focused on testing the effectiveness of robotic interventions and the development of a training regime to assist children with ASD. The present study was built on this pilot project to identify the outcomes and evaluate the essential components needed to maximise the effectiveness of the intervention programme.

This study aimed to answer the following research questions.

RQ1. What are the important qualitative outcomes that could result from robotic intervention programmes for the social development of children with ASD?

RQ2. What elements in the process of robotic intervention programmes are perceived to be conducive to the social development of children with ASD?

Methods

This study adopted a case study approach to answer the above research questions. This approach is preferable to others when ‘how’ or ‘why’ questions are posed (Yin, 2017). It allows investigators to retain in their analysis the holistic and meaningful characteristics of robotic intervention programmes that are conducted in real-life contexts. This study adopted a multiple embedded case study design. Multiple case design follows a replication logic. The cases serve in a similar manner to multiple experiments. In embedded design, each individual case study may include the collection and analysis of quantitative data (Yin, 2017). Multiple data collection methods were used, namely semi-structured interviews, document review and assessments. The unit of analysis for RQ1 was children with ASD who had participated in a robotic intervention programme, and that for RQ2 was the robotic intervention programme.

Selecting cases

Sixteen children with ASD, aged 5–11 years, were purposefully selected from mainstream schools to ensure variety in their backgrounds. Children attending mainstream school are without intellectual disability and verbally fluent. The inclusion criteria for participation in the robotic training programme, and thus this study, were based on the diagnostic criteria for ASD as stated in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, namely deficits in social reciprocity, communication and restricted and repetitive behaviours (American Psychiatric Association, 2013). Children with ASD were included if they (1) had been diagnosed with ASD and had had their diagnosis confirmed by a registered psychologist or medical practitioner, (2) had no hearing or vision deficits, (3) were able to follow simple commands, (4) were aged between 5 and 11 years and (5) did not exhibit self-injury or aggressive behaviour. Ethical approval was obtained from the Human Research Ethics Committee of the University (Ref. 2017-2018-0131).

Crafting instruments

A case study protocol was developed to guide and ensure consistency in data collection. Consistency had to be ensured because multiple data collection methods, namely semi-structured interviews, observations and standardised assessments using the Social Responsiveness Scale (SRS) and Canadian Occupational Performance Measure (COPM), were used. The case study protocol comprised an interview guide and a data collection template to guide the collection of qualitative data. Face-to-face interviews were conducted with parents and instructors to understand the impact of the programme on children and to gain their view and opinions regarding the programme.

A training guide outlining the operation of all training sessions was developed. The content of the training programme included the teaching of two-way communication, basic emotions, imitation and reciprocal responses. Throughout the twelve sessions, the participants participated in the following activities: structured interactive social games, structured story-based activities and structured singing/dancing activities.

A NAO humanoid robot was used to assist the instructor in teaching social skills to the children with ASD. NAO is designed to incorporate a wide range of behaviours, such as walking, standing up and sitting down, dancing and recognising speech, sounds and objects (Anzalone et al., 2014). The human instructor led the progression of all activities in all sessions using a non-directive approach. The robot was programmed to assist the human instructor in conducting the activities. The sessions were conducted in Cantonese. The content of the activities facilitated by the robot had been translated into Cantonese and well-tested in our pilot studies (Chung, 2020). The pilot study provided the research team opportunities to become familiarised with the operation of the robot, to conduct activities for children with ASD in school settings, and to systematically code and analyse the social behaviours of children with ASD.

The SRS is a questionnaire comprising 65 items (Constantino, 2021). The Chinese version was used in the initial screening. The SRS systematically collects information on parents’ reports of their children’s social awareness, cognition, communication, motivation and mannerisms. The questionnaire was validated and shown to have high reliability (McMahon, Vismara, & Solomon, 2013). The Chinese version was found to have good internal consistency (Cronbach’s alphas = 0.044–0.947), test–retest reliability (intraclass correlations = 0.751–0.852) and convergent validity (Gau et al., 2013).

The COPM is a validated tool for measuring intervention programmes’ outcomes related to activity participation (Law et al., 1990). It collects both quantitative and qualitative data. The COPM adopts a participatory approach, such that the participant has to articulate their expected performance outcomes, assign their priorities and rate the quality of their performance and the level of their satisfaction. Performance scores and satisfaction scores are collected both before and after the intervention to show changes in the participant throughout the intervention. It was confirmed to be a good outcome measure for use in paediatric clinical trials, as the internal consistency and responsiveness to change were high (Cusick et al., 2007).

Entering the field

Initial assessment

Before holding the intervention programme, we conducted a face-to-face interview with both the children with ASD and their parents to understand the impact of ASD on the social participation of the children at home, at school and in their wider community. The parents were asked to complete (1) the SRS to gauge their children’s social functioning and ASD symptoms and (2) the COPM to identify the essential areas of concern in the social development and participation of their children with ASD.

Robotic intervention programme

The participants received 12 weekly sessions on social skills training. The programme was individual-based and the duration of each session was 30 minutes. All sessions were conducted by a human instructor and a humanoid robot in community-based settings. The training procedures were conducted following the training protocol. Two instructors were recruited and trained to conduct the training programme using the robot. The background of these instructors was graduates in the fields of special education and rehabilitation. They were taught with the instruction techniques stated in the training protocol. They were trained and required to write observation notes for each participant in all sessions. Upon completion of the training programme, a face-to-face interview was conducted by the investigator with the instructors to explore their observations, understanding and perception of the programme.

Post-intervention assessment

Upon completion of the training programme, another face-to-face interview with both the parents and the children was conducted to explore the two parties’ perception of the feasibility and efficacy of the training programme. The COPM was re-administered.

Analysing data

The COPM

Both the qualitative and quantitative data collected by the COPM were analysed. For the quantitative component, the total scores obtained from the pre- and post-intervention assessments were calculated. The changes in both the performance and satisfaction scores were analysed using descriptive statistical analysis. A Wilcoxon signed-rank test was used to test if the changes in the scores across the two time points were significant. For the qualitative component, the reported performance issues in social development at both time points were analysed. Content analysis of all data related to all cases was conducted to identify the important outcomes related to social development, using the three essential diagnostic criteria of ASD, namely communicative behaviour, socio-emotional reciprocity and social relationship.

Interviews

Qualitative data obtained from the interviews with the parents and instructors were analysed to assess the impact of the robotic intervention programme on the social participation of the children. All of the interviews were recorded and transcribed. Important observations in the notes made by the instructors during the training sessions were recorded in the case study report. Content analysis of all data was conducted using NVivo 12, qualitative data analysis software developed by QSR International Pty Ltd. The data collected were linked to the propositions via a case description table that showed the profiles of the children.

Ensuring trustworthiness

Essential measures were taken to ensure the quality of this study. Multiple sources of evidence were used to ensure construct validity. Pattern matching and triangulation of data was used in data analysis to ensure internal validity. External validity of this study was enhanced by using of replication logic with a multiple case design. The use of case study protocol ensured the reliability of this study (Yin, 2017).

Results

Participants’ characteristics

Sixteen children with ASD completed the programme. The age of these children ranged from 6 to 10.8 years, with a mean of 7.9 years (standard deviation = 1.32). In total, 15 children were male and 1 was female. All of the children attended mainstream schools. At enrolment, the overall SRS score of these 16 children ranged from 68T to 105T, with a mean score of 79T and a standard deviation of 9.76. These pre-intervention scores demonstrated a moderate-to-severe deficit in social communication and reciprocal social behaviours that interfered with the children’s social interaction and participations. The two subscale scores, which were compatible with the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition), revealed these participants were of moderate-to-severe level of social communication (mean 76.6, standard deviation 11.49) and restricted interests and repetitive behaviour (mean 75.37, standard deviation 13.84).

COPM

The COPM results identified the successful outcomes related to social participation. The unit of analysis in this part of the study is the outcome of robotic intervention programme. A total of 48 items, which measured social development outcomes, were collected from the 16 participants. Both the performance and satisfaction scores were not normally distributed (p < 0.05); therefore, a non-parametric test was used, and the median were used to summarise the variable. The Wilcoxon signed-rank test showed that the robotic intervention programme brought statistically significant changes to the scores for both performance (Z = −5.755, p < 0.001, r = 0.59) and satisfaction (Z = −5.811, p < 0.001, r = 0.59), with actual power over 0.9. The median scores for performance were 3.5 in the pretest and 7 in the posttest. The median scores for satisfaction were three in the pretest and six in the posttest.

Content analysis was performed on the transcripts of the COPM interviews. Forty-eight successful outcomes relating to the social development of children with ASD were extracted and collated into themes and subthemes. The 12 collated themes were matched against the three essential diagnostic criteria of ASD, namely communicative behaviour, socio-emotional reciprocity and social relationship (American Psychiatric Association, 2013). Table 1 shows the results. The most commonly reported changes as revealed in the matching of codes against the themes, were verbal expression, social awareness and emotional reciprocity, in order of priority (Figure 1).

Table 1.

Thematic map showing the successful outcomes of robotic intervention in social development.

Sub-themes	Codes
Theme 1: communicative behaviour
Verbal expression	Enhanced conversation skills
	Appropriate use of verbal language to express oneself
	Use of correct grammar in spontaneous speech
	Expression on details of events
	Expression of feelings
	Reduced volume of speech to an appropriate level
	Longer, relevant and coherent verbal sentences
	Less preoccupied with self-interested topics
Non-verbal communication	Reduced inappropriate body touch with others
	Reduced inappropriate self-muttering and voice-making
	Maintained appropriate social distance with others
	Improved eye contact
	Improved manner
	Enhanced attending and listening during communication
Theme 2: Socio-emotional reciprocity
Social awareness and understanding	Less pre-occupied in self-interested topics
Social awareness and understanding	Reduced self-absorption in his own world
Reciprocal interaction	Enhanced two-ways interaction in communication
Presence and focusing	Enhanced social awareness
Presence and focusing	Enhanced concentration and focusing during interaction
Perspective taking	Social comprehension and flexibility
Perspective taking	Flexibility in making judgement
Emotional reciprocity	Give appropriate emotional response
	Improved emotional management and reduced acting out
Theme 3: Relationship
Challenging and argumentative	Reduced verbal aggression, arguing and defences
Challenging and argumentative	Less picking on others
Competition and rivalry	Less jealous competition
Competition and rivalry	Accept losing in games/competitions
Adaptive relating to others	Have more friends
Adaptive relating to others	Make appropriate new acquaintance with peers
Self-confidence and esteem	Enhanced self-confidence in trying new things and expressing oneself
Obsessions	Less obsessions with gadgets

Figure 1.

Commonly reported changes of children with autism spectrum disorder in robotic intervention.

Semi-structured interviews

Semi-structured interviews with the parents and instructors were conducted before, during and after the training programme. Content analysis of the interviews with the parents and instructors revealed the key themes regarding (1) the impact of the robotic intervention programme on the personal growth and social development of the children with ASD and (2) the elements of the programme perceived to be conducive to the growth and development of children with ASD.

Impact on personal growth and social development

Theme 1: Enhanced self-esteem

The parents thought that the interaction of their children with the robot brought a kind of enlightenment to their children. Some parents reported a positive change in the social behaviour of their children. They even received positive feedback from school teachers. They thought that their children’s encounter with the robot had enhanced the children’s self-awareness and self-image.

He is brighter than before. I guess, he thought the robot is a bit more childish than him. He said he talked with the robot and the robot could understand him. He thought he was better than the robot and therefore his esteem was boosted. (Parent B)

I could not figure out the reasons for his change. He had received only your robotic intervention programme in this period. I have never seen him behave so well and receive so much praise from others. Even his school teacher said his social behaviours had improved recently, with fewer disturbing behaviours in class. (Parent A)

Theme 2: Enhanced confidence and presence when interacting with others

The parents reported an improvement in their children’s self-confidence when interacting with others. This improvement was inferred from the children’s improved eye contact, improved assertiveness in asking questions and improved sharing of their interests with their parents, siblings and peers. The parents thought that this change might have been brought about by the improvement of their children’s basic social skills after the children attended the training programme. They said their children were happy to share their good and successful experiences communicating with the robot.

He had never shared with me regarding what he was doing while he was at school. He told me what he did with the robots after your training session. I began to prompt him to share with me his school life, though his expression was not fluent, he was able to share with me some of the routines and details. (Parent E)

I deeply felt that my son was really excited when interacting and playing with your robot. He said the robot told stories and talked with him. He said he could understand the conversations with the robot, and he could answer all questions asked by the robot. He always says he can communicate very well with the robot. To him, it is like a kind of great magic and attraction. He thinks of the robot all day and night. (Parent D)

Theme 3: Enhanced expression of emotion-laden experiences

During the intervention, some parents reported their children as expressing their experiences of classroom bullying, being misunderstood and being overstressed at school or in other community settings. They said it was not common for their children to express such emotion-laden experiences. Some children expressed such experiences to the instructors during the robotic training sessions. Some other children attempted to express their distress to their parents. However, their communication skills limited their description of most incidents, and they might have feared expressing such negative experiences. Still, when the children were asked to tell the robot about their life experiences, they were eager to describe and share their feelings and expressions, even though their verbal skills were still limited.

He expressed he was being shouted at in class by his classmates. (Parent E)

My boy said he didn’t want to go to school because he was punished by the teacher. He said he was feeling discomfort in his toe because the shoe was tight. His teacher thought that he was just irritated for no reason. (Parent F)

Elements of the programme perceived to be conducive to the growth and development of children with ASD

Theme 1: Robot friendship

Most parents and instructors reported witnessing a friendship being formed between the children and the robot. They regarded this friendship as a strong motivating factor for the children with ASD.

He likes talking to your robot. He told me the robot was eager to respond to him. He was excited and happy. There was a magic power between the robot and him. The robot is a personal friend to him. He thinks and talks about the robot all day. (Parent H)

He treated the robot as his friend. When not seeing the robot for some time, he said he missed the robot and wanted to talk with the robot soon. (Parent E)

He pretty much liked the interaction with the robot. He said he was so happy to play with the robot. I think the robot is like a friend and a playmate to him. He therefore liked the training and tried his best to learn and give good responses to his robot friend. (Instructor A)

The children really treated the robot as their good friend. Child A put tremendous effort to communicate with the robot. He told the robot about his experience in school. Child B asked the robot if it had encountered bullying at school. She even told her experience of being bullied to the robot. (Instructor B)

Theme 2: The robot as a facilitator and model

The instructors agreed that the robot was a good model from which the children could learn social behaviours and conversation skills. Given the closed-loop features and multi-modal function of the NAO robot, the robot could be an effective catalyst for interaction and facilitator in the training process.

The robot is a good role model to demonstrate the social skills somehow. It could prompt the children to respond in an appropriate way. Although the robot was talking in a monotone, he pronounced words clearly and slowly. With only simple conversation and gesture, the children were able to comprehend even the abstract ideas and complex stories. (Instructor B)

My child told me he learned from the robot that he could take a deep breath when he was being criticised by others. I didn’t imagine that he could learn from the robot. (Parent C)

Theme 3: The human instructor as a coach

The instructors reviewed the training process and found the role of the human instructor as a coach to be a critical factor contributing to the success of the programme.

To me, the robot is just a toy. The training process required my planning and leading indeed. It was difficult at the beginning. As the training course proceeded, I realised I had to act as a coach to operate the robot and guide the child to get involved in the process. (Instructor A)

Discussion

This study moved beyond laboratory studies and studied the effects of robotic intervention in a real-life context. Studies have confirmed the effectiveness of robotic intervention programmes in improving quantitative outcomes such as the frequency of eye contact, verbal initiation and joint attention (Chevalier et al., 2020; Chung, 2020). Therefore, qualitative outcomes and impact of robotic intervention on the social development of children with ASD were investigated in this study. The results from the COPM analysis showed a statistically significant change in both the levels of performance and satisfaction regarding the children’s social participation. The parents satisfaction with the performance of their children was enhanced. These results showed that the improvement in the children’s social participation was substantial, impressive and observable to their parents. The results of this study confirmed the impact of robotic intervention programmes on the daily living and social participation of children with ASD. It demonstrated that the improvement of children with ASD resulting from robotic intervention programmes had moved beyond improvement in component skills (such as eye contact and joint attention) to that in social participation in daily living.

Impressive qualitative and behavioural changes were observed and reported throughout the training. The most commonly reported changes, in order of priority, were improved verbal expression, social awareness and emotional reciprocity. The results of the thematic analyses provide a qualitative description of the children’s behavioural changes as reported by their parents and observed daily (Table 1). The children’s improvement in verbal expression was described as enhanced conversation skills, appropriate use of verbal language, use of correct grammar in spontaneous speech and giving longer and more coherent verbal responses. Improvement in social awareness was described as less preoccupation with self-interested topics and less self-absorption when interacting with others. Enhanced emotional reciprocity was described as giving appropriate emotional responses, improved emotional management and reduced acting out when relating with others. These descriptions facilitate the understanding of the qualitative and behavioural changes of the children as they encountered the social robot in a real-life context.

The personal growth and social development of the children with ASD were enhanced. Three important impacts of the programme on the children with ASD were enhanced self-esteem, self-confidence and presence. This finding is interesting. The programme was not meant to build self-esteem or self-confidence. Its primary objectives were to teach social skills. This result implies that robotic intervention programmes are not limited to enhancing social skills. It sheds lights on the implications of enhancing social development for children with ASD through their interaction and relation with a social robot.

The programme elements that were perceived as important in contributing to the positive changes in the children’s social development were (1) robot friendship, (2) the robot acting as a facilitator and model and (3) the presence of a human instructor capable of leading the programme. Robot friendship was the most commonly reported element. The forming of child–robot friendships was observed and frequently reported in this study. The formation process mostly matched that described by (Jecker, 2021): (1) the child believed and expected the robot to be a friend; (2) the child expressed feeling towards the robot, such as caring, liking and the desire to be with it and (3) the child regarded the relationship as meaningful and valuable. Even though the children in our study were all told that the NAO robot was not a human, it was observed and reported by the instructors and parents that the children treated the robot as a person and a friend. Although the children knew cognitively that the robot was not a human, they treated it as a friend and loved to play with it and enjoyed their acquaintance with it. The parents even reported that their children missed the robot.

These findings have important implications for the design of robotic intervention programmes. Robot friendship is a term used to describe the emotional connections between humans and robots. It can be in the form of companionship, emotional support and even love. Social robots are designed to have social interactions and interpersonal relations with human beings. This is made possible by essential features in the robots that enable interpersonal exchange. These features are a recognisable human or animal form, the ability to mimic human expressions and communication and social responsiveness via a closed-loop system (Ryland, 2021). Even though robot friendship is a simulated kind of friendship, through continuous dyadic interactions between the child and the robot it can develop social-emotional functions that serve to fulfil the child’s basic social needs (Bauminger et al., 2008).

There are potential drawbacks in robot friendship as developed in the robotic intervention programmes. Children may deceive themselves about the nature of a robot (Borenstein & Pearson, 2013). Children in our study perceived the robot had human qualities. However, a humanoid robot currently does not have these qualities. The moral issue may arise with robotic intervention programmes as the children may be lured into a false relationship that could bring damage to their emotional and psychological well-being (Sharkey & Sharkey, 2010). Interacting with robots may not provide the same level of social interaction and communication as human relationship. Complex social skills, such as empathy, active listening and conflict resolution, are not able to be demonstrated by robots.

Strengths and limitations were identified in this study. Regarding strengths, this study highlights the qualitative change of children with ASD in the aspects of social participation and social development. It is a systematic and robust study to adopt a case study methodology to study the phenomena in a real context. It captured important evidence of the benefits that children with ASD may gain through interactions with a social robot. It highlights the implications of robot friendship for children with ASD. However, it may be controversial to allow the children to establish robot friendships. It has been argued that human–robot friendships are ethically wrong because they are deceptive (De Graaf, 2016). The line is hard to draw. Our study supports the notion that befriending robots could actually improve the extent to which some individuals relate with and include other human beings in their social world. It is therefore suggested that the ethical issues regarding robot friendship be further explored, addressed and handled properly in future studies. Moreover, the mechanism explaining the positive change as brought by robot friendship may be worth studying in future by neuropsychologists.

Conclusion

The encounter with a social robot was regarded as meaningful and important to the children with ASD. Qualitative changes in the children with ASD undergoing a robotic intervention programme was identified in this study. The qualitative changes in social participation and personal growth through the robotic intervention programme were deemed important even to the parents of the children with ASD. However, given the ethical issues related to robot friendship, further studies may be required to clarify the conceptualisation, operation, efficacy and effectiveness of such interventions.

Key Findings

Qualitative changes in robotic intervention for children with autism were identified as enhanced self-esteem, emotional expression and social participation.

Essential elements of positive changes in robotic intervention were robot friendship, robot as facilitator and human instructor as trainer.

What the study has added

Meaning of robotic intervention to children with autism was revealed in a real context. Essential elements in robotic intervention programme conducive to positive change was identified for guiding practice.

Footnotes

Acknowledgements

The research team would like to thank all participants in their contribution in this study.

Research ethics

Ethical approval from the Human Research Ethics Committee of The Education University of Hong Kong was obtained (2017-2018-0131).

Consent

All participants were given written informed consent for participation in this study.

Patient and public involvement data

During the development, progress and reporting of the submitted research, patient and public involvement was not included at any stage of the research.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the General Research Fund (18609719) from the Research Grants Council, University Grants Committee, HKSAR.

Contributorship

EC contributed to develop the theoretical framework, research design, coordinate data collection, preform data analysis and interpret results, as well as the writing of the manuscript. DC and KS participated in the checking of data and interpretation of results and drafting of the manuscript.

ORCID iD

Eva Yin-han Chung

References

American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders (DSM-5®). Washington, DC: American psychiatric association.

Anzalone

Tilmont

Boucenna

, et al. (2014) How children with autism spectrum disorder behave and explore the 4-dimensional (spatial 3D time) environment during a joint attention induction task with a robot. Research in Autism Spectrum Disorders 8: 814–826.

Bauminger

Solomon

Aviezer

, et al. (2008) Children with autism and their friends: A multidimensional study of friendship in high-functioning autism spectrum disorder. Journal of Abnormal Child Psychology 36: 135–150.

Borenstein

Pearson

(2013) Companion robots and the emotional development of children. Law, Innovation and Technology 5: 172–189.

Chevalier

Kompatsiari

Ciardo

, et al. (2020) Examining joint attention with the use of humanoid robots-A new approach to study fundamental mechanisms of social cognition. Psychonomic Bulletin & Review 27: 217–236.

Chung

(2020) Robot-mediated social skill intervention programme for children with autism spectrum disorder: An ABA time-series study. International Journal of Social Robotics 13: 1095–1107.

Costa

Lehmann

Dautenhahn

, et al. (2015) Using a humanoid robot to elicit body awareness and appropriate physical interaction in children with autism. International Journal of Social Robotics 7: 265–278.

Constantino

(2021) Social Responsiveness Scale. In: Volkmar

(Ed.), Encyclopedia of Autism Spectrum Disorders. Springer International Publishing, pp. 4457–4467.

Cusick

Lannin

Lowe

(2007) Adapting the canadian occupational performance measure for use in a paediatric clinical trial. Disability and Rehabilitation 29(10): 761–766.

10.

Dalemans

de Witte

Lemmens

, et al. (2008) Measures for rating social participation in people with aphasia: A systematic review. Clinical Rehabilitation 22: 542–555.

11.

De Graaf

(2016) An ethical evaluation of human–robot relationships. International Journal of Social Robotics 8: 589–598.

12.

Diehl

Schmitt

Villano

, et al. (2012) The clinical use of robots for individuals with autism spectrum disorders: A critical review. Research in Autism Spectrum Disorders 6: 249–262.

13.

Fazli

Imani

Abedini

(2018) Faculty members’ experience of student ethical problems: A qualitative research with a phenomenological approach. Electronic Journal of General Medicine 15(3): em23.

14.

Gau

Liu

, et al. (2013) Psychometric properties of the Chinese version of the social responsiveness scale. Research in Autism Spectrum Disorders 7(2): 349–360.

15.

Giannopulu

Pradel

(2010) Multimodal interactions in free game play of children with autism and a mobile toy robot. NeuroRehabilitation 27: 305–311.

16.

Huijnen

Lexis

Jansens

, et al. (2019) Roles, strengths and challenges of using robots in interventions for children with autism spectrum disorder (ASD). Journal of autism and developmental disorders 49: 11–21.

17.

Jecker

(2021) My friend, the robot: An argument for e-friendship. In: 2021 30th IEEE International Conference on Robot & Human Interactive Communication (RO-MAN). IEEE, pp. 692–697.

18.

Law

Baptiste

McColl

, et al. (1990) The Canadian occupational performance measure: An outcome measure for occupational therapy. Canadian Journal of Occupational Therapy 57: 82–87.

19.

McMahon

Vismara

Solomon

(2013) Measuring changes in social behavior during a social skills intervention for higher-functioning children and adolescents with autism spectrum disorder. Journal of Autism and Developmental Disorders 43: 1843–1856.

20.

Robins

Dautenhahn

Te Boekhorst

, et al. (2004) Effects of repeated exposure to a humanoid robot on children with autism. In: Designing a more inclusive world. London: Springer, pp. 225–236.

21.

Ryland

(2021) It’s friendship, Jim, but not as we know it: A degrees-of-friendship view of human–robot friendships. Minds and Machines 31: 377–393.

22.

Shamsuddin

Yussof

Ismail

, et al. (2012) Initial response in HRI-a case study on evaluation of child with autism spectrum disorders interacting with a humanoid robot Nao. Procedia Engineering 41: 1448–1455.

23.

Sharkey

(2010) The crying shame of robot nannies: An ethical appraisal. Interaction Studies: Social Behaviour and Communication in Biological and Artificial Systems 11(2): 161–190.

24.

Simut

Vanderfaeillie

Peca

, et al. (2016) Children with autism spectrum disorders make a fruit salad with probo, the social robot: An interaction study. Journal of autism and developmental disorders 46: 113–126.

25.

Srinivasan

Lynch

Bubela

, et al. (2013) Effect of interactions between a child and a robot on the imitation and praxis performance of typically developing children and a child with autism: A preliminary study. Perceptual & Motor Skills 116: 885–904.

26.

Yin

(2017) Case study research and applications: Design and methods. USA: SAGE.