Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

ROBOTICS AS A DIDACTIC TOOL FOR STUDENTS

WITH AUTISM SPECTRUM DISORDERS: A

SYSTEMATIC REVIEW

La robótica como herramienta didáctica para personas con desórdenes en el

espectro del autismo: una revisión sistemática

Itsaso Arocena Perez

itsaso.arocena@ehu.eus

https://orcid.org/0000-0002-1551-4405

University of the Basque Country (Spain)

Asier Huegun Burgos

asier.huegun@ehu.eus

https://orcid.org/0000-0003-0703-0766

University of the Basque Country (Spain)

Itziar Rekalde Rodríguez

itziar.rekalde@ehu.eus

https://orcid.org/0000-0002-9297-6734

University of the Basque Country (Spain)

Recibido: 01/02/2021

Evaluado: 04/03/2021

Revisado: 06/05/2021

Aceptado: 24/05/2021

Resumen

Este artículo describe los resultados cuantitativos y cualitativos de un estudio

cuyo objetivo es identificar las tendencias y oportunidades de innovación en el

campo de la robótica socioeducativa, utilizada como herramienta didáctica, con

el fin de desarrollar las diferentes habilidades, destrezas y competencias de los

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

estudiantes con necesidades específicas de apoyo educativo escolarizados;

estudiantes con trastornos del espectro del autismo específicamente. Para ello

se ha realizado una revisión sistemática de la literatura, mediante una

estrategia de búsqueda rigurosamente definida. Los resultados obtenidos

permiten identificar los avances en cuanto a modelos didácticos basados en el

uso de la robótica como herramienta educativa, actividades pedagógicas y

recursos didácticos; criterios, estrategias e instrumentos de evaluación y

experiencias de aplicación en contextos escolares reales

Abstract

This article describes the quantitative and qualitative results of a study aimed to

identify the trends and innovation opportunities in the social robotics area. Using

robots as a didactic tool to develop different skills, abilities, and competencies of

students with specific educational support needs who attend mainstream school

classes; students with autism spectrum disorders specifically. A systematic

review of the literature was carried out through a rigorously defined search

strategy. The results obtained allow us to identify the advances in didactic

models based on robotics as an educational tool, pedagogical activities and

didactic resources, evaluation criteria, strategies and instruments, and

application experiences in real school contexts.

Palabras Clave: Trastornos del espectro del autismo, escuela, robótica socio-

educativa

Keywords: Autism spectrum disorders, school, social robotics

Introduction

Social robotics represents an innovative area, especially in education (Bekele,

Crittendon, Swanson, Sarkar & Warren, 2014). During the last years, articles

written about social robotics have multiplied (Billard, Robins, Nadel &

Dautenhahn, 2007). Day by day, more researchers begin to see robots as

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

valuable tools in educational processes, especially for those with special

educational needs.

In the case of people with autism specifically, many experiences show us how

they are attracted to social robots, how they like and enjoy spending time with

them (Fortis, Goedert & Barrett, 2011; Dunst, Trivette, Prior, Hamby & Embler,

2013; Huskens, Palmen, Van der Werff, Lourens & Barakova. 2015). However,

do robots give them something positive besides fun? Do they improve their

quality of life in any way?

When we review the literature about social robotics written lately, we can see

that it is not easy to find special educators or teachers in the primary databases

on education (Web of Science, Scopus, and ERIC). Most of the articles are

written from computer engineering or mechanical engineering. That is, they

have not been addressed from an educational point of view.

Autism spectrum disorders nowadays

Which are the main theories about ASD?

There are several explanatory theories of autism spectrum disorders. These are

the ones that are commonly accepted by the scientific community (Wellman,

2016).

Theory of mind

The theory of mind allows us to modulate our social behavior, makes us able to

put ourselves in the place of the other, and makes us understand our emotional

circumstances (Yun, Choi, Park, Bong & Yoo, 2017).

We decipher what is in the other's mind using the information given by their

facial expression, body posture, and intonation to express themselves

(Baron‐Cohen, Campbell, Karmiloff‐Smith, Grant & Walker, 1995). The more we

know a person, the more we know about their knowledge, intentions, beliefs,

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

and desires, and the easier it gets for us to know what is "going through their

mind." In ASD, the theory of mind deficits carries difficulties, such as

understanding one's mind and others' minds, understanding the rules that

regulate social interaction, communicative use of language, and symbolic game

(Baron-Cohen et al., 1995).

Central coherence theory

Central coherence theory claims that people on the spectrum of autism tend to

process information in high-level units of meaning, losing details (Valdez, 2001).

On the one hand, this information processing involves a more remarkable ability

for analytical tasks and better processing of the visual versus the linguistic. On

the other hand, less ability in global stimulus processing, difficulties in

differentiating relevant information from irrelevant information, and problems

processing the information in a contextualized way (Martos-Pérez, Llorente-

Comí, 2013).

Executive function theory

The executive functions are necessary mental abilities. They give us context to

execute our mental functions efficiently (Willcutt, Doyle, Nigg, Faraone &

Pennington, 2005). We can say that they unify, order, and coordinate all other

cognitive functions. Deficits in executive function explain the rigidity people on

the spectrum experiment when facing changes, problems to define goals and

design strategies, difficulties in anticipating the future, presence of repetitive,

stereotyped behavior, and restricted interests (Hujinen, Lexis, Jansens & de

Witte, 2016).

Extreme male brain theory

This theory was proposed by Simon Baron-Cohen (2002). Although it is

becoming increasingly unpopular, extreme male brain theory argues that there

are differences between male and female brains. ASD people's brains could act

as they have a male brain taken to the extreme. The observable characteristics

would be more systematic thinking, more detail fixing, less empathic capacities,

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

less social abilities, and less skilled conversation (Warren, Zheng, Das, Young,

Swanson, Weitlauf & Sarkar, 2015).

Why is it important to develop joint-attention in children with ASD?

People who grow up and progress ordinarily develop joint attention mainly

through symbolic play and effective interaction with their caregivers. First joint

attention signs begin to occur between 9 and 18 months of age (Charman,

2003). On the contrary, this phenomenon does not occur in people who are

within the autistic spectrum. Unlike children with typical development, children

with ASD are often more involved with their thoughts and feelings than with

other people (Carrasco, Alarcón & Trianes, 2018; Charman, 2003; Matsuda,

Nunez, Hirokawa, Yamamoto & Suzuki, 2017).

Joint attention consists of a group of nonverbal behaviors that include directing

the gaze, pointing, and teaching objects, which refer to external stimuli during a

communicative exchange (Nowell, Watson, Faldowski & Baranek, 2018), as

well as the ability to keep focused on something simultaneously with our

partners (Carrasco et al., 2018).

This ability allows us to cooperate and socialize with others. Interacting with

other children during childhood is fundamental to developing the social and

language skills we need in adulthood (Matsuda et al., 2017).

In addition, as seen in Figure 1, we know that joint attention is a pivotal ability,

so to say, a skill that is fundamental for developing other areas of functioning.

Because of this, the development of pivotal abilities produces generalized

behavioral improvements in children (Charman, 2003; Weiss and Harris, 2016).

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 1. Structure of pivotal abilities

Note. Prepared by the author based on Charman, 2003; Nowell et al. 2018; Matsuda et al.,

2017

We can conclude that by developing these essential or key skills, we will

naturally improve all others, thus working on a generalized development of the

quality of life of the person with ASD and facilitating their daily routines.

Which are the common interventions to develop joint-attention in primary and

secondary school?

Below are different interventions that are currently used to improve the quality of

life of people with ASD (Fuentes-Biggi, Ferrari-Arroyo, Boada-Muñoz, Touriño-

Aguilera, Artigas-Pallarés, Belinchón-Carmona & Díez-Cuervo, 2006). There

are many more, but we have chosen those based on scientific evidence.

The Discrete Trial Training (DTT) method breaks down specific skills in small

steps, which are learned gradually, so they are taught from attention skills to

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

more complex ones such as verbal skills or social behaviors. It starts from

simple skills, increasing complexity as the child progresses (Mulas, Ros-

Cervera, Millá, Etchepareborda, Abad & Téllez de Meneses, 2010). The

methodology is based on four elements:

First of all, the therapist presents a stimulus as a precise order or question. If

necessary, the order is followed by reinforcement. Secondly, the child responds

correctly or incorrectly. Finally, the therapist provides a consequence: a correct

answer receives a boost while an incorrect one is ignored or corrected.

The Denver model presents a checklist composed of four levels of objectives

divided into months old the child is (Kim, Berkovits, Bernier, Leyzberg, Shic,

Paul & Scassellati, 2013). The model divides the development of the child into

four levels: 12 to 18 months, 18 to 24 months, 24 to 36 months, and 36 to 48

months. These goals are milestones that children with ASD need to improve to

achieve a better quality of life in their early years. This curriculum was designed

from the observation of babies with ASD symptoms to reinforce those most

challenging areas. The evaluation is done together with the parents and is

mainly observational (Rogers & Dawson, 2010).

Pivotal Response Training (PRT) is a variation of Applied Behavioral Analysis

(ABA) type therapy. It focuses on pivotal areas: increasing a child's motivation

to learn, initiating communication, joint attention, and self-regulation. By

focusing on these main areas, the effects of treatment carry over into many

aspects of a child's behavior and skills, including social, communicative, and

academic skills (Minjarez, Mercier, Williams & Hardan, 2013).

PRT incorporates into the child's everyday routine, along with family members,

peers, teachers, and other professionals (Forment-Dasca, 2017). Each program

is carefully customized to the needs of a specific child. The essential

components of PRT include:

• Treatment takes place in the natural environment of the kid: home or

school

• Family involvement

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

• Coordination with professionals across all environments (teachers,

educators)

• Treatment focused on pivotal areas.

The SCERTS model plans objectives for the child with ASD to meet for a

comprehensive evaluation of this in all its contexts. The SAP (SCERTS

Assessment Process) is a model evaluation process made of two parts; the

recording one and the observational one. It is a long and extensive process and

allows educators to determine the key objectives in the intervention. It is

composed of three stages.

The first one is the social partner stage. The child communicates with their

peers through presymbolic communication systems, gestures, or vocalizations.

The second one is the linguistic couple stage, where the child communicates

with peers by single words to combinations of several words besides early

symbolic communication systems. Finally, the conversational couple stage,

where the child communicates with their peers through phrases and interactive

speech.

Is it possible to use social robotics as educational intervention tools?

Understanding and using social skills is the most challenging developmental

area for people with autism spectrum disorders (Weiss and Harris, 2016). They

feel, live, and express their affection and emotions in different and personal

ways, making communicative exchanges complicated, which can cause

frustration on the participants of said communicative actions.

During the last decades, various types of educational interventions using

animals have emerged, especially for people with functional diversity. The value

of these interventions assisted by animals is commonly accepted; contact with

an animal positively influences the self-perception of the human being and

stimulates their social behavior. This effect is called a social catalyst since it

facilitates interpersonal interactions (Beetz, Uvnäs-Moberg, Julius & Kotrschal,

2012). Dealing with animals makes people with ASD trust others more easily,

develop greater empathy and increase their self-esteem, as they feel able to do

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

something difficult for them (Simut, Vanderfaeillie, Peca, Van de Perre &

Vanderborght, 2016).

The animals that have proven to be most effective are dogs. However, we can

also find studies conducted with horses, cats, and other animals, which show

beneficial effects of this type of intervention in people with ASD (Silva, Correia,

Lima, Magalhães & de Sousa, 2011).

The most significant difference between using animals and using robots as a

social catalyst lies in controlling the situation. Pets can be trained to be great

therapists, but animals will always have a series of unexpected behaviors facing

a situation, which we can not completely control. However, we can fully program

a robot and adapt it to respond to the needs that arise during an interaction.

This control gives us flexibility and a more remarkable ability to reach all users,

regardless of their circumstances (Beetz et al., 2012).

Recent studies show that people with ASD tend to feel comfortable interacting

with social robots due to their low emotional stimulation (Huijnen, Lexis,

Jansens & de Witte, 2016; Kumazaki, Warren, Swanson, Yoshikawa,

Matsumoto, Ishiguro & Kikuchi, 2018).

Interacting with robots can be particularly enriching for a child with autism

spectrum disorder, as it can overcome the barriers experienced in face-to-face

interaction with other people. However, there is always a person behind the

robot who must design the objective of the intervention and all the appropriate

didactic sequence to achieve it (Hashim and Yussof 2017; Huijnen et al., 2016).

Aims of the current study

The purpose of this study is to provide a systematic review of the scientific

literature about using robotics as tools in educational intervention with people

with ASD in order to fulfill the following goals:

• Determine the number of articles that have been published in ERIC,

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

WOS, and Scopus databases

• Identify which is the scientific field where research is being done on this

topic

• Determine the impact factor of the articles that tackle this topic

• Analyze the chronological evolution of this topic in the current scientific

production

• Identify the main objectives to be developed through robotics

• Determine if the objectives proposed are fulfilled

• Describe the scenario where the interventions have been made

• Establish the participant number on the interventions

• Identify the duration of the intervention

• Detect the main robot models used and their characteristics

Method

This study aims to deepen the field of science that combines social robotics and

learning difficulties, specifically autism spectrum disorders, from an educational

perspective. The most relevant databases for research in the socio-educational

field were chosen; Web of Science (WOS), Scopus, and Education Resources

Information Center (ERIC).

The method chosen was a systematic review. Kitchenham Brereton, Budgen,

Turner, Bailey, & Linkman (2009) claim: "Researchers performing a systematic

review must make every effort to identify and report research that does not

support their preferred research hypothesis as well as identifying and reporting

research that supports it" (p.5).

The keywords chosen were autism and robotics. Using the thesaurus the

database provides us with, we could do more exhaustive research using the

following keywords; On the one hand, autism, Asperger syndrome, and

behavior disorders. On the other hand, social robotics, cybernetics, and

electronics.

Considering that our focus and interest are on primary education and secondary

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

education, all other educational levels were excluded, leaving us with the final

figure of 28 studies in ERIC, ten studies in WOS, and seven studies in Scopus.

Figure 2. Visualization of the research process

The analysis started through the ERIC database because it references

education, continued with WOS and ended up with Scopus. As the work

progressed, we noticed that some articles were repeated in the three databases.

*These repeated articles were manually discarded. We are aware that the

number of articles found in this search will vary from when the article was

written until its publication date since socio-educational robotics is a field that is

now emerging.

In order to keep these articles organized, an index file was created (Díaz-

Posada, Varela-Londoño & Rodríguez-Burgos, 2017). The selected articles

were sorted according to the date, from the oldest to the newest. In addition, the

following data were also collected in the files; the name of the article, journal,

type of publication, country of publication, and name and surname of the

authors. Those files were encrypted with code names and added to other files

according to the items analyzed.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Results

Firstly, the scientific field where the papers belong was analyzed. To do so, the

areas from which the authors of the selected articles came were taken into

account. Most of them show us the result of collaborative works, so they belong

to different fields simultaneously.

Figure 3. Scientific field

Engineering is the field from which more was researched on the educational

robotics subject (Figure 3), closely followed by education. Finally, something

less has been investigated from psychology and medicine.

Afterward, we will focus on the years in which these articles were published.

Figure 4. Publication year

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

The first article found in this search was published in 1968. Since then, there

have only been one or no publications per year, until we arrived in 2013, where

there is a prominent production peak (Figure 4). As of this moment, production

declines although it remains higher than in the previous period.

Figure 5

Objectives

Hence objectives of the articles chosen were analyzed. They were classified

into three areas; Acquire knowledge, develop abilities and compare situations.

Watching Figure 5 gives us the feeling that it is commonly accepted that social

robotics can be used to improve different skills in people on the autism

spectrum.

Most of the works aims to develop the skills of the participant. Many of them

seek to gain knowledge in this area, and some aim to compare different

situations. Each area was individually analyzed.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 6

Objective: Acquire knowlidge

Henceforth, we will focus on those articles that aim to gain knowledge (Figure 6).

The majority does it through understanding the cybernetic systems and

explaining how they work. Some works aim to review the subject

bibliographically, and the rest aims to explain or publicize the working

methodology used.

Figure 7

Objevtive: Develop abilities

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

As Figure 7 shows, most of the works intended to increase the vocal production

of the participants in terms of skill development. In the second place, we find

those works aimed to increase the communicative skills of the participants, and

finally, we find those works aimed at developing joint-attention and social skills

to the same extent.

Figure 8

Objective: Comparison

The vast majority of comparative works base their comparison on the same

scenario; the same intervention is repeated with both a human and a robot

(Figure 8). Only one of the analyzed works compares the effectiveness of the

robot with a traditional toy.

Now we turn from the objectives to the conclusions of the articles analyzed.

Regarding the level of compliance with the objectives set, ten of the papers met

their objectives, six of the works claim not to have met their objectives, and two

of them show an intermediate level of compliance.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Table 1

Objectives met

Objectives not

met

Unexpected

objectives

Not educational

objectives

As seen in Table 1, 33,3% of the articles have undefined educational objectives.

Those who do not come from pedagogy have the object of studying the robot,

its design, and its operation. Therefore, after doing a couple of sessions with

children, the conclusions obtained throw no light on robotics as an effective tool

to work with people on the spectrum of autism. In this section, we want to clarify

that we refer to those articles that do not have any goal regarding the students

as not educational objectives.

On the one hand, among the works that claimed to meet their objectives, six

articles aimed to communicate with the robot. Two of them intended to improve

the students' vocal production, and two others made comparisons of two similar

situations. In one of them, students interacted with the robot, and in the other,

students interacted with a person.

On the other hand, we have articles that did not meet their objectives. Four of

these works compared two similar situations. In one, the student interacted with

the robot, whereas in the other one, the student interacted with a human.

Another one of the articles that did not meet their objectives tried to increase the

students' vocal production through a robot. The last of the articles aimed to

compare students' behavior after performing an intervention of a few sessions.

However, when comparing these results with the control group, which did not

participate in the intervention, no conclusive differences were found.

To conclude with this section, we will mention two articles that achieved

objectives and unachieved objectives. The first of these aimed to reduce the

anxiety of the participants and improve their social skills. After finishing the

intervention, it was observed that although students did reduce their anxiety,

there was no significant improvement in social skills.

Finally, we found an article that the authors present as unfavorable, although it

could be considered differently. This work aimed to ensure that students imitate

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

the robot better than a person. It was carried out with a group of people on the

autism spectrum and neurotypical people. Interestingly, people on the spectrum

imitated the robot better, while neurotypical students imitated the person better.

Next, the works in which socio-educational interventions were carried out were

analyzed. Specifically, we focused on whether these interventions were carried

out in the participants' natural environment or occurred in an artificial

environment.

Figure 9

Scenario

As we can see in Figure 9, most of the interventions were carried out in artificial

environments, such as; a research center or an isolated room inside a school.

Few were made on natural scenarios to the participants, as their school,

ordinary classroom, or even their own homes.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 10

Participant number

Figure 10 shows that more than half of the interventions analyzed were carried

out with less than ten participants, and one-third of the studies analyzed were

put through with less than five participants in them.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 11

Inclusion criteria

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

All the articles were analyzed to explain selecting the participants and the

sample of participants in detail. However, only six articles mention the selection

criteria used: they were based on choosing the intervention participants.

In Figure 11, we can see the inclusion criteria that these researchers used when

choosing participants for their interventions. As shown in the image, the

condition to participate in the most repeated interventions is to have a minimum

intelligence quotient.

The second most repeated criterion is very close to the first and is the vocal

production; Some researchers require their participants a minimum of spoken

production capacity to participate in their project.

The third most repeated request is the age of the participants. All studies ask for

participants between the ages of four years and 13 years. We conclude from

this that the literature focuses on students who are in the primary school stage.

Fourth, we have another criterion that refers to the skills of the potential

participants. Some researchers ask students who will participate in their

interventions not to have any visual or hearing disability, nor should they have

any psychiatric disorder. We should remember that we are analyzing those

works that do include the selection criteria. Since most of the articles analyzed

in this meta-analysis do not mention any criteria, we can find students with

Down syndrome or generalized developmental disorder.

We start now with those criteria that are mentioned once. The first one says that

the selected participants should not be participating simultaneously in another

intervention. Another one of the selection criteria says that the participating

students should show a lack of social initiation.

Continuing with the skills required of the participants, one of the studies requires

its potential participants to have the ability to recognize emotions and facial

expressions. We have left for the end the selection criteria that have seemed

most relevant to us, and that is inclusion in schools. One of the articles

stipulates that participant must attend school and must be included in the

scholar system.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 12

Intervention duration

The next item analyzed was the duration of the interventions. 59,7% of the

interventions examined lasted less than one month, 27.3% lasted between one

and three months. Finally, a single work carried out an intervention beyond

three months (Figure 12).

In the next section, the methodologies used in the reviewed research will be

analyzed. As we can see, most of the interventions have used video recordings

to collect their data. Out of 23 articles that have used the video recording

methodology, 21 have defined their indicators of success and have analyzed

the recordings quantitatively, adding numerical values to each segment of the

video, according to the number of indicators met. However, two of these 23

articles have analyzed the videos quantitatively but using pre-existing research

applications to do so.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 13

Research methodology

Following the numerical order, in second place, we find the theoretical works,

three of which are intervention proposals. Five are reviews of previously carried

out interventions, making a total amount of eight papers. In third place are the

investigations that have used scales to measure their results. They used pre-

existing scales, three used the test system, and three used the pre-test and

post-test system.

Finally, we find the least used data collection methodologies in this field: the

focus group and the interview. A Focus group was used in two investigations to

collect opinions and perceptions of the participants. The interview was used in a

single intervention, along with a scale, for the same purpose.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 14

Robot models

As we can see in Figure 14, the most used robot among the analyzed works

was Nao, closely followed by Popchilla, Kaspar, and Probo. All the works

carried out to explain the characteristics of the robot that they chose to carry out

the intervention. However, only four of them justify this choice.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Figure 15

Robot model justification

As we can observe in Figure 15, the characteristics that make researchers

choose a robot model or another are mostly programmability, accessibility, the

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

humanoid form, or the shape of a stuffed toy. Depending on the objectives of

the intervention, one or the other form was chosen.

Discussion

We can arrange the difficulties that a person in the spectrum of autism lives in

three major categories; Difficulties of social interaction, communication

difficulties, and difficulties of activity and interest (Taylor, 2015; Javed & Park,

2019). Of these three blocks, we can see that the literature focuses more on

working on the first two.

However, it is known that working on pivotal skills like joint attention, other areas

such as communicative language or empathy improve as well (Charron & Craig

2017). It seems logical to think that if individuals feel more comfortable in their

social environment, they will feel less anxiety. Their communicative language

will improve as well as their vocal production. They will be more inclined to

communicate with the people around them.

Nevertheless, most of the works carried out were aimed to improve vocal

production, that is, getting the child to say more words by using a robot.

Although vocal production is essential, it is even more critical to work joint

attention and social skills with people on the spectrum of autism (Hujinen, Lexis,

Jansens & de Witte, 2017), which, as a curiosity, make up the less researched

objectives in this area.

The individual with ASD condition likes things to stay the same, with no changes.

Often, the person with ASD has difficulty moving from one thing to another

(Jung, Lee, Cherniak & Cho, 2019). In the case of children, this is shown as

stress when a transition is necessary (Taylor, 2017). Considering this, we can

expect that bringing the child to an experimental room where he/she has not

been before, his/her behavior will not be the same as it usually is.

It will take a while for them to accustom to this new situation. We will have to

establish a routine where we carry out the same sessions simultaneously for a

while so that the results are conclusive. Moreover, the goal of a robot should be

to provide educational support in the classroom, never to replace the teacher

(Warren, Zheng, Swanson, Bekele, Zhang, Crittendon & Sarkar, 2015).

Despite this information, the vast majority of the researches have not taken

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

place in the natural environment of the kids (home, school), but they were

carried out in experimental rooms. In addition, in many cases, they made less

than five sessions total, so we consider that the results shown by these

experiments are inconclusive.

In the fifth edition of its diagnosis and statistical manual of mental disorders, the

American Psychiatric Association calls the condition so far known as ASD, more

precisely, disorders of the autism spectrum (American Psychiatric Association,

2013). To be accurate, we could not even talk about disorder, and we would talk

about people within the spectrum of autism.

We call it that because it is an extensive range, where each person is different

from the rest, they have their strengths and weaknesses (Silva et al., 2011). We

say that someone is within this spectrum when it picks up some common

characteristics such as those mentioned up to now. For all this, we consider that

an investigation with less than 5 participants in it cannot be considered

transferable to other environments, nor can generalizable conclusions be drawn

from it.

Finally, we were surprised that only one of the articles analyzed mentioned

selection criteria to recruit participants. The children selected must be attending

school and receive a formal education (Kim et al., 2013). As educators, we

believe in educational inclusion and the importance of each student's space in

the classroom.

Conclusion

The total number of articles found in the three databases on educational

robotics with people with disorders in the autism spectrum has been 39. As this

is a current and booming issue, we expected to find many more but were not so.

This literature sprout could be because we sought primary education and

secondary education among the keywords and robotics since we wanted to see

how robotics was being used in schools.

Regarding the scientific field from which the analyzed interventions were carried

out, we can see that, firstly, most of the selected studies have been carried out

from engineering. Secondly, we found the studies carried out by professionals in

education, as educators or special education teachers. Finally, a minority

percentage of the selected articles were made from psychology and medicine.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

We can see then that it is in engineering and education where there is greater

interest in this subject.

Regarding the impact factor, we must say that, on the one hand, only slightly

more than half of the articles analyzed have an impact factor on Journal Citation

Report. On the other hand, each journal belongs to a different scientific area,

with its criteria when measuring impact. Taking into account both factors, we will

not venture to conclude this point.

Chronologically speaking, we can observe a very high production peak in 2013.

Since then, it dropped a bit, but a constant production was maintained. It seems

to be a new topic, which is unknown about much, but the scientific community

trusts.

In terms of objectives, the majority of the analyzed articles aimed to increase

the vocal production of the participants, followed closely by those articles that

aimed to improve the communicative skills of the participants. Considering that

communicating with their environment is one of the most significant difficulties

people with autism face in their daily lives, looking for these capabilities seems

coherent.

On the one hand, it seems that the use of robotics in people on the spectrum of

autism is a subject that arouses interest in the scientific community since there

are many works whose objective is to delve into existing knowledge. On the

other hand, there are also many works whose objective focuses on the robot;

this means that the child is not the protagonist but a way to test the robot.

Finally, we see the same amount of work that explains a work methodology or

an intervention. As educators, we think that the representation of research

projects with educational purposes is shallow.

Regarding the level of fulfillment of these objectives, most articles claimed to

have fulfilled their purpose. This majority is just over half, though, which leaves

us almost half of the articles analyzed, which could not meet the objectives set

on their interventions.

Next, we will put our attention on the scenario where the selected interventions

were put through. Almost all of the interventions were carried out in a fictitious

scenario, created for the intervention, not in the participant's natural scenario,

such as students' classroom or even students' own home. This scenario

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

election does not seem appropriate since people with autism suffer when

changes and uncertain situations happen while they feel comfortable within tight

spaces and routines.

Almost half of the studies carried out were accomplished with less than ten

participants. In addition, almost all of the selected works lasted less than a

month. We believe that more accurate data would be obtained by performing a

more extensive intervention and more participants. In addition, the inclusion

criteria used when choosing the participants of the intervention are not

substantiated. Some articles mention some required characteristics when

choosing their participants, such as age or intellectual quotient. However, there

is no justification for why these attributes make these participants more

appropriate for such interventions.

Let us look at the criteria used to choose the subjects of the interventions. We

can see that a part of the scientific community considers robots more

appropriate or beneficial in people on the spectrum of autism but require less

support.

Finally, we will focus on the robot model used in the selected interventions. The

most used robot is Nao, followed by Popchilla, Kaspar, and Probo. Except for

Nao, which was created by the French company Alderaban, the rest of the

robots were created by research teams from different universities. We can

conclude that there is no agreement on the scientific community on the

characteristics that make a robot better than the others to work with people on

the spectrum of autism.

Although robotics is not a recent issue, its application in the field of education is.

We live in the boom of social and educational robotics right now; it is time to join

forces and investigate where this path leads us.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of

mental disorders (5th ed.). Arlington, VA.

Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in

cognitive sciences, 6(6), 248-254.

Baron‐Cohen, S., Campbell, R., Karmiloff‐Smith, A., Grant, J., & Walker, J.

(1995). Are children with autism blind to the mentalistic significance of the

eyes?. British Journal of Developmental Psychology, 13(4), 379-398.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Beetz, A., Uvnäs-Moberg, K., Julius, H., & Kotrschal, K. (2012) Psychosocial

and psychophysiological effects of human-animal interactions: the

possible role of oxytocin. Front. Psychology 3:234.

Bekele, E., Crittendon, J. A., Swanson, A., Sarkar, N., & Warren, Z. E. (2014).

Pilot clinical application of an adaptive robotic system for young children

with autism. Autism, 18(5), 598-608.

Billard, A., Robins, B., Nadel, J., & Dautenhahn, K. (2007) Building robota, a

mini-humanoid robot for the rehabilitation of children with autism.

Assistive Technology, 19(1), 37-49.

Carrasco, C., Alarcón, R., & Trianes, M. V. (2018). Adaptación y trabajo

cooperativo en el alumnado de educación primaria desde la percepción

del profesorado y la familia. Revista de Psicodidáctica, 23(1), 56-62.

Charman, T. (2003). Why is joint attention a pivotal skill in autism? Philosophical

Transactions of the Royal Society B: Biological Sciences, 358(1430),

315–324

Charron, N., Lewis, L., & Craig, M. (2017). A robotic therapy case study:

Developing joint attention skills with a student on the autism spectrum.

Journal of Educational Technology Systems, 46(1), 137-148.

Díaz-Posada, L. E., Varela-Londoño, S. P., & Rodríguez-Burgos, L. P. (2017).

Inteligencias múltiples e implementación del currículo: avances,

tendencias y oportunidades. Revista de Psicodidáctica, 22(1), 69-83.

Dunst, C. J., Trivette, C. M., Prior, J., Hamby, D. W., & Embler, D. (2013).

Parents' Judgments of the Acceptability and Importance of Socially

Interactive Robots for Intervening with Young Children with Disabilities.

Social Robots Research Reports, Number 1. Orelena Hawks Puckett

Institute.

Forment-Dasca, C. (2017). Modelos de intervención en los trastornos del

espectro autista: Denver y SCERTS. Revista Neurologia, 64, S33-S37.

Fortis, P., Goedert, K. M., & Barrett, A. M. (2011). Prism adaptation differently

affects motor-intentional and perceptual-attentional biases in healthy

individuals. Neuropsychologia, 49(9), 2718-2727.

Fuentes-Biggi, J., Ferrari-Arroyo, M. J., Boada-Muñoz, L., Touriño-Aguilera, E.,

Artigas-Pallarés, J., Belinchón-Carmona, M., & Díez-Cuervo, A. (2006).

Guía de buena práctica para el tratamiento de los trastornos del espectro

autista. Rev neurol, 43(7), 425-38.

Hashim, R., & Yussof, H. (2017). Humanizing humanoids towards social

inclusiveness for children with autism. Procedia Computer Science,

105(4), 359-364.

Huijnen, C. A., Lexis, M. A., Jansens, R., & de Witte, L. P. (2016). Mapping

robots to therapy and educational objectives for children with autism

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

spectrum disorder. Journal of autism and developmental disorders, 46(6),

2100-2114.

Huijnen, C. A., Lexis, M. A., Jansens, R., & de Witte, L. P. (2017). How to

implement robots in interventions for children with autism? A co-creation

study involving people with autism, parents and professionals. Journal of

autism and developmental disorders, 47(10), 3079-3096.

Huskens, B., Palmen, A., Van der Werff, M., Lourens, T., & Barakova, E. (2015).

Improving collaborative play between children with autism spectrum

disorders and their siblings: The effectiveness of a robot-mediated

intervention based on Lego® therapy. Journal of autism and

developmental disorders, 45(11), 3746-3755.

Javed, H., & Park, C. H. (2019). Interactions With an Empathetic Agent

Regulating Emotions and Improving Engagement in Autism. Ieee

Robotics & Automation Magazine, 26(2), 40-48.

Jung, S. E., Lee, K., Cherniak, S., & Cho, E. (2019). Non-sequential Learning in

a Robotics Class: Insights from the Engagement of a Child with Autism

Spectrum Disorder. Technology, Knowledge and Learning,

10.1007/s10758-018-9394-8

Kim, E. S., Berkovits, L. D., Bernier, E. P., Leyzberg, D., Shic, F., Paul, R., &

Scassellati, B. (2013). Social robots as embedded reinforcers of social

behavior in children with autism. Journal of autism and developmental

disorders, 43(5), 1038-1049.

Kitchenham, B., Brereton, O. P., Budgen, D., Turner, M., Bailey, J., & Linkman,

S. (2009). Systematic literature reviews in software engineering–a

systematic literature review. Information and software technology, 51(1),

7-15.

Kumazaki, H., Warren, Z., Swanson, A., Yoshikawa, Y., Matsumoto, Y., Ishiguro,

H., & Kikuchi, M. (2018). Impressions of humanness for android robot

may represent an endophenotype for autism spectrum disorders. Journal

of autism and developmental disorders, 48(2), 632-634.

Kumazaki, H., Yoshikawa, Y., Yoshimura, Y., Ikeda, T., Hasegawa, C., Saito, D.

N., & Matsumoto, Y., (2018). The impact of robotic intervention on joint

attention in children with autism spectrum disorders. Molecular autism,

9(1), 46.

Martos-Pérez, J., & Llorente-Comí, M., (2013). Tratamiento de los trastornos del

espectro autista: unión entre la comprensión y la práctica basada en la

evidencia. Revista de neurología, 57(Supl 1), S185-91.

Matsuda, S., Nunez, E., Hirokawa, M., Yamamoto, J., & Suzuki, K. (2017).

Facilitating social play for children with PDDs: Effects of paired robotic

devices. Frontiers in psychology, 8, 1029.

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

Minjarez, M. B., Mercier, E. M., Williams, S. E., & Hardan, A. Y., (2013). Impact

of pivotal response training group therapy on stress and empowerment in

parents of children with autism. Journal of Positive Behavior Interventions,

15(2), 71-78.

Mulas, F., Ros-Cervera, G., Millá, M. G., Etchepareborda, M. C., Abad, L., &

Téllez de Meneses, M., (2010). Modelos de intervención en niños con

autismo. Rev Neurol, 50(3), 77-84.

Nowell, S. W., Watson, L. R., Faldowski, R. A., & Baranek, G. T. (2018). An

initial psychometric evaluation of the joint attention protocol. Journal of

autism and developmental disorders, 48(6), 1932-1944.

Rogers, S. J., & Dawson, G. (2010). Early Start Denver Model for young

children with autism. New York, NY: Guilford.

Silva, K., Correia, R., Lima, M., Magalhães, A., & de Sousa, L. (2011). Can dogs

prime autistic children for therapy? Evidence from a single case study.

The journal of alternative and complementary medicine, 17(7), 655-659.

Simut, R. E., Vanderfaeillie, J., Peca, A., Van de Perre, G., & Vanderborght, B.

(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(1), 113-126.

Taylor, G. (2015). Trastornos del Espectro Autista Guía Básica para Educadores,

Madrid, España, Narcea Ediciones.

Valdez, D. (2001). Teoría de la mente y espectro autista. D. Valdez. Autismo.

Enfoques actuales para padres y profesionales de la salud y de la

educación. Buenos Aires: Fundec.

Warren, Z. E., Zheng, Z., Swanson, A. R., Bekele, E., Zhang, L., Crittendon, J.

A., & Sarkar, N. (2015). Can robotic interaction improve joint attention

skills? Journal of autism and developmental disorders, 45(11), 3726-3734.

Warren, Z., Zheng, Z., Das, S., Young, E. M., Swanson, A., Weitlauf, A., &

Sarkar, N. (2015). Brief Report: Development of a Robotic Intervention

Platform for Young Children with ASD. Journal of Autism and

Developmental Disorders, 45(12), 3870-3876. 10.1007/s10803-014-

2334-0

Weiss, M. J., & Harris, S. L. (2001). Teaching social skills to people with autism.

Behavior modification, 25(5), 785-802.

Wellman, H. (2016). Social cognition and education: theory of mind.

Pensamiento educativo. Revista de Investigación Educacional

Latinoamericana, 53, 1-23.

Willcutt, E. G., Doyle, A. E., Nigg, J. T., Faraone, S. V., & Pennington, B. F.,

(2005). Validity of the executive function theory of attention-

Revista científica electrónica de Educación y Comunicación en la Sociedad del Conocimiento

Publicación en línea (Semestral) Granada (España) Época II Vol. 21 (1) Enero-Junio de 2021 ISSN: 1695-324X

https://revistaseug.ugr.es/index.php/eticanet

DOI: http://doi.org/10.30827/eticanet.v21i1.18137

deficit/hyperactivity disorder: a meta-analytic review. Biological psychiatry,

57(11), 1336-1346.

Yun, S., Choi, J., Park, S., Bong, G., & Yoo, H. (2017). Social Skills Training for

Children with Autism Spectrum Disorder Using a Robotic Behavioral

Intervention System. Autism Research, 10(7), 1306-1323.

10.1002/aur.1778