Introduction
Literature Review
Parents of Children with ASD
Parents in Traditional Autism Therapy
Parents in Robot-Assisted Autism Therapy
Current Study
Methods
Participants
Child | Age | Verbal | ADHD | ADOS-2 | N | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
C1 | 5–6 | – | – | 8 | 6 | P | N | N | P | N | P | ||||
C2 | 5–6 | – | – | 5 | 6 | P | N | P | N | N | N | ||||
C3 | 8–9 | – | – | 8 | 6 | P | N | N | N | N | N | ||||
C4 | 5–6 | ✓ | – | 4 | 5 | N | N | P | N | N | |||||
C5 | 5–6 | ✓ | ✓ | 6 | 5 | P | N | P | N | N | |||||
C6 | 7–8 | – | – | 7 | 4 | P | P | N | P | ||||||
C7 | 8–9 | ✓ | – | 9 | 2 | P | N | ||||||||
C8 | 10–11 | ✓ | ✓ | 9 | 9 | P | P | N | N | N | N | N | N | N | |
C9 | 7–8 | – | ✓ | 9 | 9 | P | N | P | P | P | P | P | P | P | |
C10 | 7–8 | – | ✓ | 8 | 8 | P | P | P | P | N | N | N | N | ||
C11 | 5–6 | – | ✓ | 6 | 8 | P | N | N | N | P | P | P | P | ||
C12 | 5–6 | – | ✓ | 7 | 8 | P | N | N | N | N | N | P | P | ||
C13 | 6–7 | – | – | 8 | 7 | P | P | P | P | P | N | P | |||
C14 | 11–12 | – | – | 9 | 6 | P | P | P | N | N | N | ||||
C15 | 9–10 | ✓ | ✓ | 8 | 6 | P | N | N | N | N | N | ||||
C16 | 5–6 | – | ✓ | 6 | 7 | P | P | P | N | N | P | P |
Robot
Activities
-
Positive reinforcement was added throughout all robot activities to encourage children during the intervention. Children with autism need more stimulation and reward from the environment to practice behaviors and succeed in using them in the future. The robot was programmed to perform a number of verbal and non-verbal reinforcement behaviors. The examples of verbal praise include “Well done”, “Keep up the good work”, “Perfect”, and others. While non-verbal behaviors were the following: clapping hands, cheering, smiling, raising arms, and others. For instance, for the Touch me activity, we programmed the robot to clap hands accompanied by the corresponding sound when the child identified the correct body part and pressed the appropriate tactile sensor (e.g. left foot’s bumper). Additionally, at the end of each session, children were given a selection of stickers to choose from.
-
Picture exchange communication systems (PECS) were utilized in several activities including Emotions, Transport, Animals and Storytelling. For example, for each emotion in the Emotions activity, the robot pointed at the photo with a situation (e.g. waiting on the bus stop), performed animation and told what emotion it felt (e.g. “I feel bored waiting for the bus”), and then pointed at the photo with a child showing that emotion (e.g. a boy being bored). Children’s joint attention was practiced in this activity. Similarly, the printed images of transports, animals, and storytelling characters were used as PECS in other activities.
-
Errorless teaching is also applicable to our intervention framework. For example, in the “Touch Me” application, there was no differentiation between right and left arms and legs but only between arms, legs, and head tactile sensors. Also, when a child did not know how to touch the robot`s sensor correctly, a human therapist showed how to proceed.
-
Focusing on peer-mediated social skills training (Battaglia & Radley, 2014), we programmed the robot’s role as a peer with the help of verbal utterances such as “let's play together”, “we are going to play”, and others. Additionally, the use of pronouns were included throughout all verbal utterances such as “we/ours” and informal “you/your” (Russian translation to “ты/твoй’’ and Kazakh translation to “ceн/ceнiң”). This is in fact demonstrated in the survey on the use of the NAO robot, i.e. children often refer to the NAO as a friend or a peer (Amirova et al., 2021).
Setup
The Role of the Therapist and Researchers
Two Session Conditions: With and Without Parents
Procedure
Video Coding
Measures | Description | Adopted from |
---|---|---|
Engagement | Mean of engagement scores calculated for each session: a 5-point Likert scale, where 1—full non-compliance, 2—non-compliance, 3—several prompts, 4—one/two prompts, and 5—immediate reaction | Kim et al., (2012) Pop et al., (2014) |
Valence | Mean of valence scores for each session: a 5-point scale: 1—cry/anger/fear, 2—sad/bored, 3—neutral, 4—interested, and 5—happy/excited | Kim et al., (2012) Rudovic et al., (2017) |
Engagement time | The amount of time a child was engaged during one session. This variable is calculated relative to the overall time of the session (e.g. Engagement time is 25% i.e. 3 min out of 12 min-session) | Kim et al., (2012) Rudovic et al., (2017) |
Eye gaze time | The amount of time a child spent looking at the robot calculated relative to the overall duration of the session | Admoni et al., (2011) Pop et al., (2014) |
Affection | The duration of actions (kissing, hugging, tender touching, scratching, petting) of a child that are calculated relative to the overall duration of the session | Stanton et al., (2008) |
Curiosity | The frequency of actions (opening, rotating, touching body parts) calculated relative to the overall duration of the session | Stanton et al., (2008) Rudovic et al., (2017) |
Aggression | The frequency of actions (pushing, biting, hitting, pulling fingers) calculated relative to the overall duration of the session | Stanton et al., (2008) Pop et al., (2014) |
Chest button | The total number of pressing a chest button for each session | |
Stereotyped behaviours | The amount of time a child was flapping their hands, screaming, and crying were calculated relative to the overall time of the session | Stanton et al., (2008) Pop et al., (2014) |
Smiles | The total number of times a child smiles in each session. Each second was coded as 1 for smiles and 0 if not detected | Pop et al., (2014) Rudovic et al., (2017) |
Words | The total number of spoken words during a session linked to the time | Stanton et al., (2008) |
Thematic Analysis of Parents’ Interviews
Results
Parental Presence for All
Children with Severe ASD
ASD with Co-occurring ADHD
Verbal and Non-verbal Children
Interviews with Parents
Discussion
Parental Presence
Severe Autism
Verbal Abilities
Parent Observations
Key Takeaways on Parental Involvement in the RAAT
-
(1) Parental input to the therapy helps identify behavioral characteristics and personal preferences of children with ASD if researchers consult with parents before conducting the interventions. The initial consultation with parents allows for the design of better treatment options.
-
(2) Parents should be aware of their roles and responsibilities during therapy to minimize the unwarranted intervention.
-
(3) For severe conditions of autism, parents could be present in the initial (few) sessions to allow their children to accustom to the new environment, robot and/or the therapist.
-
(4) There might still be some children who need their parents around them at all times, and their whole experience with the robot would be impossible without their parents` presence.
-
(5) Robots serve mediating roles and may stimulate special bonding between children and their parents or co-present others. Children tend to share their novel experiences with the robot with people closest to them.
-
(6) Parents’ attitudes towards target activities may be collected through qualitative (e.g., interviews, observation) as well as quantitative (e.g. surveys) data collection methods. This would ensure the triangulation and transferability of results. Caregivers can identify and report minimal changes in their children’s behaviors and act as reliable informants. At the same time, their responses have to be interpreted with caution as there could be perception biases in parents.
-
(7) Parents should be encouraged to take time for themselves. Some parents may stop attending the sessions after realizing that their children feel good and comfortable with the robot. They therefore can take that free time to recharge.
-
(8) In collaboration with therapists, researchers may develop guidelines or tips for parents to use behavioral strategies such as positive reinforcement and prompting. These skills are helpful both within and outside therapy.