1 Introduction

The beginning of schooling is a time for the child’s adaptation to a new environment, which will require communication, academic and social skills that are important for insertion into school. The first stage of Basic Education, Early Childhood Education, seeks to ensure “accessibility of spaces, materials, objects, toys and instructions for children with disabilities, global developmental disorders and the gifted” (Resolution no. 1, 1999, p. 19).

Thus, children diagnosed with Autistic Spectrum Disorder (ASD), enrolled in Early Childhood Education, need their needs to be met, support and services to be guaranteed. According to the proposals of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), the deficit in social communication is one of the domains of diagnostic criteria of ASD, being characterized as a decrease in the ability to use verbal and non-verbal communication during an interaction with other people, in addition to limitations in sharing ideas and interests or negotiating in a positive and friendly way (American Psychological Association [APA], 2014).

Given these needs and the demands that are required by society, specifically in the process of schooling in Early Childhood Education, the intervention in Supplementary and/or Alternative Communication is a possibility to be implemented in the school to support the communicative limitations of children who have complex communication needs. Thus, an evidence-based practice in Supplemental and/or Alternative Communication is the Picture Exchange Communication System (PECS), developed in a training protocol format by ^{Bondy and Frost (1994)} while working with children in the Delaware Autism Program in 1985.

The authors applied PECS training to 85 children in Early Childhood Education, noting that 85% of children learned to make requests using at least two pictures. The researchers pointed out that 76% of the children who used PECS started using speech concomitantly. Only in 1994, the authors developed the PECS training manual in order to guide professionals and parents in understanding the approach and apply it in a safer and more reliable way. Initially, the program was applied to children with ASD who could not use speech as a primary means of communication. Currently, it has been applied to people of different ages and diagnoses to increase their communication skills within the social context.

From the international literature referring to the PECS, it was verified in three meta-analysis studies (^{Flippin, Ruzka, & Watson, 2010}; ^{Hart & Banda, 2010}; ^{Tincani & Devis, 2011)}, covering the years 1994 to 2009, that of the 19 studies analyzed that applied the PECS only two managed to complete the six phases of the program. In the systematic review of national literature, carried out by ^{Mizael and Aiello (2013)}, the starting date was not defined because there was no systematic review on the subject; however, the year 2011 was set as the deadline when the study ended. Thus, the authors found six studies, but none applied the six phases of the PECS. Tincani and Devis (2011) point out some factors in relation to the non-application of all phases of the PECS and the delay in its completion. One of them was the criterion of phase change of 80% of correct answers, which could prevent researchers from completing all phases of the system until the conclusion of the research. Another limitation refers to the participants’ cognitive skills that could have influenced learning, this being a hypothesis, since most studies did not have enough descriptions of the participants’ cognitive functioning levels for this skill to be considered a variable that influenced the acquisition of the PECS. However, empirical data from Phases V and VI of the system are scarce.

Literature has shown that children with ASD are characterized by severe attention deficits, in addition to not being prone to learning skills by observing the behavior of others. This deficit puts them at an intense disadvantage when trying to maintain the educational, social and functional pace of their peers. Even if the acquisition of an imitative repertoire is proposed by explicit instruction, children with ASD do not necessarily acquire an observational learning repertoire (^{Bandura, 1977}; ^{Greer, Dudek-Singer, & Gautreaux, 2006}; ^{Plavnick & Hume, 2014)}.

In this perspective, based on ^{Bandura’s theory (1969)}, Video Modeling (VM) is a technique that has been widely used and researched. Studies point to the effectiveness in acquiring and developing different skills for people with ASD, such as: social skills, communication skills, vocational skills, games, physical activities, everyday life activities, imitation, team training, academic and attention skills (^{Bellini & Akullian, 2007}; ^{Huang & Wheeler, 2006)}. Video Modeling is characterized by ^{Shukla-Mehta, Miller and Callahan (2010)} with the following procedure: a) a person is invited to watch the video; b) the skill to be developed is modeled by an adult or peer in an activity context; c) the instructor provides stimuli and reinforcement for the person to respond to relevant stimuli; d) the person imitates the model’s behavior with the opportunity to perform the skills shown in the video.

There are three types of VM: VM with others as a model (VMO), Video Self-Modeling (VSM) and Point-of-view Video Modeling (POVM). The VMO uses an adult or peer as a model, who the learner may or may not know; VSM is a specific form of video modeling that uses the learner as a model to offer learning opportunities, in which he/she visualizes him/herself performing a task (^{Rodrigues & Almeida, 2017)}. And finally, Point-of-view Video Modeling (POVM), which consists of footage shot from the perspective of the first person. The video is made as if the camcorder were at shoulder height and as if it were viewing the target behavior to be performed through the person’s eyes (^{Shukla-Mehta, Miller, & Callahan, 2010)}.

Given the nature of the visual perspective, external stimuli are not included in the video, naturally directing the viewer’s attention to the necessary steps to complete the desired task or demonstrate appropriate behavior. Despite the greater efficiency and focus on the task, POVM has received less empirical attention than other types of videos. POVM is highly effective for participants with ASD and developmental disabilities, although the results indicate significant effects for those with ASD. As already noted, participants with ASD are typically more responsive to visual stimuli, and interventions based on VM benefit from this preference (^{Bellini & Akullian, 2007)}. POVM has a differential aspect in relation to other videos, since irrelevant external stimuli are significantly reduced due to the nature of the footage, allowing the viewer to focus on the relevant stimuli (^{Mason, Davis, Boles, & Goodwyn, 2013)}.

Due to the effectiveness of Video modeling in teaching various skills, studies focusing on its association with PECS have shown that learning becomes faster because, when observing someone performing the target behavior in the PECS phases, the student needs less help, reaching the necessary adjustments for phase changes in less time.

The studies that carried out the association between PECS and Video modeling were performed by ^{Cihak, Smith, Cornett and Coleman (2012)}; Smith, Hand and Dowrick (2014); ^{Rodrigues, Campos and Almeida (2015)}. The study conducted by Cihak et al. (2012) evaluated the increase in independent initiatives for communication through a combination of Phase I of PECS associated with video modeling in children with autism at preschool ages. The authors concluded that students were able to learn Phase I of PECS and increased independent initiations for communication in that phase. However, students learned Phase I of PECS faster when they used it associated with VM. Smith, Hand and Dowrick (2014) examined the effectiveness of VSM for teaching PECS up to Phase IV. The participants were two boys (9 and 11 years old) with autism and a man with Down Syndrome (36 years old). All three participants were non-verbal and had no functional communication system; the two boys had a long history of PECS failure. All participants showed rapid learning of their target behaviors when VSM was introduced; moreover, the effects were generalized without the need for further intervention.

^{Rodrigues, Campos and Almeida (2015)} applied the PECS associated with the VMO to a child with Down Syndrome of nine years old, in all phases of the PECS. The results indicated that the participant’s performance was satisfactory, both in relation to each phase, and in relation to the learning of the entire intervention process. The participant learned the PECS in a few sessions and it was observed that the use of the picture was a support for the increase of vocabulary present in the formulations of words and phrases, both through the pictures and through speech. The authors demonstrated an intervention possibility that favored the participant’s communicative skills.

According to the studies outlined in the literature on this subject, this study aims to analyze the effects of the Picture Exchange Communication System (PECS) associated with Point-of-view Video Modeling (POVM) on the communication skills of children enrolled in Early Childhood Education and diagnosed with Autistic Spectrum Disorder (ASD) and Complex Communication Needs.

2 Method

The study was submitted and approved by the Research Ethics Committee of the Federal University of São Carlos (UFSCar), under the Opinion no. 45113615.4.0000.5504. Thus, the participants’ communication skills were the dependent variable and the PECS alone and the PECS associated with the POVM were the independent variables. For this purpose, it was conducted through the Multi-mode Survey Design combined with the Alternating Treatment Design (^{Gast & Ledford, 2010)}.

The Multi-mode Survey Design is similar to the Multiple Baseline Design, since the independent variable (in the case of this study, initially there were two independent variables: PECS alone and PECS + POVM) is systematically and sequentially introduced, at different times, behaviors, environments or subjects. However, it differs from the multiple baseline design with respect to the baseline that is not collected continuously in behaviors that are not under intervention conditions. On the contrary, the baseline is conducted intermittently in the behaviors that will be subjected to the intervention (^{Gast & Ledford, 2010)}.

The Alternating Treatment Design allows comparing interventions. In this study, intervention 1 (PECS alone) and intervention 2 (PECS associated with POVM) are alternated in their application to compare which intervention is most effective. There should be at least five interventions from each. After comparing the interventions, the study may be concluded once there is stability of the results or remain with the most effective intervention until its conclusion (Wolery, ^{Gast, & Hammond, 2010)}.

3 Results and discussion

The results of the participants regarding the implementation of the PECS associated with the POVM are represented in Figure 1, which shows the number of correct responses during the intermittent baseline, intervention and follow-up conditions for participants P1, P2 and P3.

Source: Research database.

Figure 1 Performance of participants. 

From the data presented in Figure 1, none of the participants obtained correct answers regarding the target behavior during the baseline sessions. From the moment that interventions with PECS alone and PECS associated with POVM began, the levels and tendencies of their data changed - all participants obtained the criterion, reaching a 100% non-overlapping. The scores show that there was an immediate and abrupt change in the dependent variable after the introduction of the independent variable. The fact that the intervention includes stimuli, tips, reinforcement organization, provides a high probability of an increase in correct responses and an abrupt level change.

When comparing Phase I, with regard to the conditions of PECS alone and PECS associated with POVM, the average percentage of correct answers was calculated for each condition. In this perspective, for all participants, PECS associated with POVM was more effective than PECS alone, since 100% of participants had a higher number of correct answers with PECS associated with POVM, with an average of 89.3% (variation: 86% to 94.2%) for PECS and 95.1% (range: 92% to 98.4%) for PECS associated with POVM. Thus, respecting the parameters of the design used, the intervention was continued with the condition that proved to be most effective: the PECS associated with the POVM.

In Phase II, based on the information in Figure 1, the PECS intervention associated with POVM remained. All participants were able to reach the goal of the phase, which was to walk to the communication folder, pick up the picture, walk to the communication partner, deliver the picture and receive the item. The phase change criterion was reached in five sessions for all participants, with an average of independent responses for the exchange of pictures from 97.2% (range: 88% to 100%) to P1, 90.6% (range : 70% to 98%) for P2, 93.4% for P3, with an average of 93.7% (range: 84% to 100%) of independent responses.

In Phase III-A, P1 and P2 needed a minimum number of sessions to reach the phase change criterion, with the average of the independent responses for both being 100% (M: 100%; variation: 100%); thus, in three consecutive sessions, they obtained the largest number of independent responses and met the criterion for phase change. For P3, it took five sessions to reach the phase change criterion; in the last three consecutive sessions, obtained 100% of independent responses (M: 95.6%; variation: 81% to 100%). In this phase, everyone managed to discriminate between the interesting item picture and the uninteresting item picture. Thus, P1 and P2, at no time, took the picture of the uninteresting item, demonstrating great ability to discriminate pictures. Although P3 needed five sessions to reach the criterion, he/she also demonstrated many hits in discriminating the pictures.

In Phase III-B, all participants were able to match the picture to the item. At that time, even though there were five pictures of interesting items, participants were able to select the picture of the item they wanted and pick up the item corresponding to the picture, demonstrating their ability to discriminate. For P1, three sessions were necessary; in all of them, it obtained 100% (M: 100%; variation: 100%), reaching the criterion for the phase change. For P2, five sessions were necessary; in the last three consecutive sessions, obtained 100% (M: 98%; variation: 93% to 100%), reaching the criterion for phase change. For P3, five sessions were held; he/she reached the criterion for phase change from the average of the sessions (M: 97.2; variation: 94% to 100%).

In Phase IV, the participants obtained an average of independent responses of 93.4% (range: 91.4% to 94.4%). For P1, four sessions were necessary and, in the last three consecutive sessions, obtained 100% (M: 94.3%; variation: 77% to 100%) of independent responses, reaching the criterion for phase change. For P2, five sessions were necessary, in which, in the last three, obtained 100% (M: 91.4%; variation: 76% to 100%) of independent responses. For P3, five sessions were carried out and, in the last three consecutive ones, obtained 100% (M: 94.4%, variation: 82% to 100%) of independent responses. At the beginning of this phase, participants demonstrate a drop in the number of independent responses, as it involves many new skills such as taking the sentence initiator “I want”, putting it on the sentence strip, picking up the picture of the item, putting it on the sentences, removing the sentence strip and giving it to the communication partner to read the message. Due to this large number of new steps, the participants needed more physical assistance to complete the sequence, but in the subsequent sessions, they learned the entire sequence and reached the criterion for the phase change.

In Phase V, the overall average of the participants was 97.2% (range: 96.2% to 97.8%). For P1, five sessions were necessary, reaching an average greater than 80% for the phase change (M: 97.6%; variation: 96% to 100%). For P2, five sessions were necessary, in which the average was greater than 80% (M: 97.8; variation: 91% to 100%). For P3, five sessions were necessary in which the average was greater than 80% (M: 96.2%; range: 92% to 100%). In this phase, all participants were able to answer, through the pictures, the question: “What do you want?”.

In Phase VI, the participants used, in addition to the phrase initiator “I want”, other initiators. This phase encourages the increase of spontaneous comments, its continuity is permanent. The overall average of the participants was 96.1% (range: 95% to 96.8%). For P1, four sessions were held and in the last three sessions obtained 100% (M: 95%; variation: 80% to 100%) of responses independently, reaching the criterion for phase change. For P2, five sessions were necessary and the average of the sessions was greater than 80% (M: 96.8%; variation: 90% to 100%). For P3, five sessions were necessary, the average of which was greater than 80% (M: 96.6%; range: 93% to 100%).

The Follow-up was collected for all participants. The data showed that everyone maintained the target behaviors (exchanging pictures to communicate independently) with 100% of independent responses one month after the end of the intervention, demonstrating that they are able to use PECS independently, without any need for assistance.

In this study, the effect of using PECS associated with Video Modeling proved to be effective in facilitating the learning of participants with ASD and complex communication needs. In Phase I, in the PECS intervention alone, the participants demonstrated 86% to 94% of correct answers, which characterizes a good performance. However, in the sessions in which PECS was associated with Video Modeling, the successes ranged from 92% to 98%, considering that this second condition favored a faster learning and with a greater number of correct responses than PECS alone.

In the subsequent phases (II to VI), considering the most satisfactory performance with the use of PECS associated with POVM in Phase I, the intervention with POVM continued. Thus, the performance of the participants remained above the baseline condition, considering that no session was below 58%. Thus, in all phases, participants reached the criterion of phase change in a maximum of five sessions, requiring an average of 43 intervention sessions to complete the six phases of the PECS. According to the international literature, only one study presents the six phases of PECS for children with ASD (^{Charlop-Christy, Carpenter, Le, LeBlanc, & Kellet, 2002)}. In the national literature, only two studies applied the six phases of PECS associated with Video Modeling (^{Rodrigues, 2015}; Rodrigues, Campos, & Almeida, 2015); however, they involved participants with Down Syndrome. In this sense, the results of the present study can confirm that POVM assisted the learning of children with ASD in all six phases, contributing to the completion of training in a short time, in a few sessions and with most attempts with independent exchanges.

These results were similar to those found in the study by ^{Cihak et al. (2012)}, in which the participants, two with autism and two with three years of development delay, had more independent initiatives in Phase I of PECS when they used it associated with Video modeling with others as a model. In the study performed by ^{Smith, Hand and Dowrick (2014)}, the authors demonstrated that the use of PECS associated with self-modeling (Video Self-Modeling) can provide an accessible and easy-to-use alternative for teaching supplementary and alternative communication for two 11-year-old boys with ASD and a 36-year-old adult with Down Syndrome. The study conducted by ^{Plavnick and Vitale (2016)} demonstrated that video modeling was more effective than live modeling in teaching verbal operants (exchanging pictures for the desired item).

A limitation of the study is related to aspects such as number of participants. In this sense, the designs of single subjects, such as the multi-mode survey design, suggest more replications. Further studies may be carried out so that the results can be compared.

Future research may replicate studies with people with ASD of the most diverse ages and diagnoses, in addition to involving other levels of education. The results of this study may enable the use of video modeling in clinical and educational practice in teaching PECS, in addition to enabling its use for teaching other strategies in supplementary and/or alternative communication, as in the study performed by ^{Mason et al. (2013)}, in which they used Video Modeling to develop visual contact skills, facial expressions, shift changes and shared emotions of two participants with autism. Considering the Brazilian reality, this intervention may be a possibility in schools, given its low cost, as well as bringing contributions to students with complex communication needs.

3.1 Social validity

The results of the social validity questionnaire (Figure 2) are presented below, which aimed to verify the degree of satisfaction and importance of the people directly involved with the participants and the study.

Source: Research database.

Figure 2 Result of the social validity questionnaire. *PP1 = Teacher of P1, PP2 = Teacher of P2, PP3 = Teacher of P3, PEE: Teacher of Special Education, RP1 = Responsible for P1, RP2 = Responsible for P2, RP3 = Responsible for P3, EP1 = P1 trainee, EP2 = P2 trainee, EP3 = P3 trainee. 

Then, some responses from the people who participated in the social validity questionnaire were selected. Those involved were asked to spontaneously talk about their perceptions about the study.

P3 teacher: “He’s smart, he knows, he understands, but sometimes he couldn’t say what he understood. Wow, I loved it so much, it was an opportunity, thankfully he had that opportunity, I´d say that helped at school, will help at home, will help in his life, very good, I thank you”.

Responsible for P3: “As there are people who think that PECS is a very repetitive thing, that the child is robotic and that will not stimulate speech, as in the case of P3, who does not speak at all, will he only stick to the picture exchange? I do not agree. It worked very well and if his mind works with images, his mind, we notice that it is a disorganized arcade; in this case, PECS helped to organize thought and he will only speak with his thought organized. (...). He distinguishes one picture from the other, this is excellent for brain work. It shows that he has a reasoning and that his vision is good, because there were people who even said that he was not blind, so I see beyond that, I see not only PECS, because even if he does not speak, he communicates perfectly through the pictures, one day he may let a word out. Like he dropped spoons, a cart, it didn’t happen anymore, but it may happen. The autistic person needs several repetitions, it is a daily routine of repetition, and in a few attempts, he succeeded, he is already achieving it faster than before, you model once, you do a modeling and it is enough. That’s why I think it works”.

Responsible for P2: “It’s really good (laughter), he goes to bed and takes the folder with him, he puts it aside, I take it out, I put it in another way, he doesn’t, it’s like that, that way there. I said, I can’t stand this folder anymore (laughter), where I go this folder is also there, here is the place for the folder, you sit somewhere else, go for it, it’s like that (laughter). If we are talking to him, anywhere, he assembles the sentence, he takes it everywhere he goes, you have no idea, if he goes to the kitchen he takes it with him. Excellent, very good”.

P2 trainee: “I think it was very important, in his communication, everything, I think he evolved a lot. At first, I thought that, for me, I thought it was going to be difficult at first, but from the first time he started, wow, it was excellent”.

P1 teacher: “I think it was very important for her development, for communication, right, her with us because until then we did not understand what she expressed, right, and now, through the folder, we already understand better what she wants to express, what she wants”.

The comments were illustrative as to the perception they had of the intervention performed, demonstrating positive aspects for participants with ASD and Complex Communication Needs: alternative communication enabled participants to demonstrate their communication skills that had not yet been manifested, increased vocabulary, support for speech intelligibility and improvement in social interaction.

4 Final considerations

The PECS associated with POVM proved to be effective in teaching communication skills to children with Autistic Spectrum Disorder and Complex Communication Needs. Therefore, the participants had a large number of independent responses, characterizing the learning of the PECS. As for communication skills, there were gains in the skills predicted by the PECS and in skills not predicted.

One aspect that deserves discussion involves the application of PECS associated with POVM in the educational context. Considering the nuances present in this environment such as: number of students, presence of a single teacher, curriculum planning to be carried out; applying an individualized program in a collective environment is a challenge. However, research that takes place in natural environments, such as school, benefits the training and learning of all those involved, in line with the notes shown by ^{Schirmer and Nunes (2011)}, considering that on-site research offers greater opportunities for development and participation of all. In this sense, it can be concluded that it is possible to apply the PECS associated with POVM in the classroom environment of regular teaching; however, many factors corroborate for it, in fact, to happen, but these factors are not related to the limitations of the PECS, but, rather, to factors specific to school dynamics.

Finally, providing means of communication is to enable survival, independence and quality of life. Supplementary and/or Alternative Communication must be guaranteed to children, adolescents and adults with Complex Communication Needs, and this guarantee must come mainly from legislation, public policies, with the encouragement of research and the disclosure of its results.