Video Self-Modeling for a Student with Dravet Syndrome: An Intervention Involving Parents during COVID-19 Pandemic in Italy

Video self-modeling instruction offers advantages compared to in-vivo instruction but has not been used with individuals with Dravet syndrome. Therefore, the purpose of this study was to investigate the effects of video self-modeling (VSM) on three different behaviors of a 12-year-old boy with Dravet syndrome. We taught the participant’s mother to use video-modeling instruction via role-playing and feedback and evaluated effects of VSM using a multiple-baseline design across behaviors: ordering numbers in descending sequence, positioning features on a face, and reading words. The VSM increased performance accuracy for all three skills, suggesting that VSM interventions via telehealth may provide an effective and sustainable option for skill development. Supplementary Information The online version contains supplementary material available at 10.1007/s43494-021-00063-1.

The COVID-19 global pandemic has changed the delivery of rehabilitation services for people with disabilities. Telehealth interventions can help contain the spread of the virus, especially for the services that involve close contact between the client and therapist, like applied behavior analysis (ABA). In Italy, ABA services for people with autism and intellectual disabilities are not recognized as essential health services. Therefore, professionals have made a transition, at least temporarily, towards telehealth services during which services are delivered through videoconferencing systems (Rodriguez, 2020). Telehealth may also be beneficial after the pandemic if it contributes to better results or provides more efficient or cost-effective outcomes.
In Italy, during the pandemic, the opportunity for in-person rehabilitation services decreased dramatically for families with children with disabilities. Therefore, parents and other family members have taken on essential roles in supporting and managing their children, and many families have not been trained in delivering specialized interventions based on ABA. Previous research has shown that video modeling, in which the behavior is demonstrated through a video, can increase intervention fidelity by paraprofessionals who have little or no training in discrete trial training (DTT) with a high level of social validity (Cardinal et al., 2017). Video models also reduce the need for a behavior analyst during initial training and reduce the cost of parents' interventions in the home setting (Barboza et al., 2019).

Abstract
Video self-modeling instruction offers advantages compared to in-vivo instruction but has not been used with individuals with Dravet syndrome. Therefore, the purpose of this study was to investigate the effects of video self-modeling (VSM) on three different behaviors of a 12-year-old boy with Dravet syndrome. We taught the participant's mother to use video-modeling instruction via role-playing and feedback and evaluated effects of VSM using a multiple-baseline design across behaviors: ordering numbers in descending sequence, positioning features on a face, and reading words. The VSM increased performance accuracy for all three skills, suggesting that VSM interventions via telehealth may provide an effective and sustainable option for skill development.
There are various video-modeling methods to teach academic skills (e.g., Kellems et al., 2016). Among these, video self-modeling (VSM) involves the individual observing a recording of themselves accurately performing the target behavior (i.e., a competent model; Mason et al., 2016). VSM has improved social communication skills (Bellini & Akullian, 2007), functional skills (Marcotte et al., 2020)., academic skills (Boon et al., 2020)., and behavioral functioning (Prater et al., 2012). for students with disabilities. To our knowledge, no studies have evaluated the effectiveness of telehealth-based VSM for participants with Dravet syndrome (DS), a rare form of childhood-onset epilepsy that is characterized by frequent and prolonged seizures, accompanied by psychomotor and neurological disorders (Dravet, 2011), which that can severely impair social life and adaptive behavior (e.g., Vascelli et al., 2021). In this study, VSM was used during parent-delivered DTT sessions for a participant with DS. The objective was to improve academic performance on three skills.

Participant and Setting
The participant, Andrew, was a 12-year-old boy with DS who attended the first class of lower secondary schools in Italy. Andrew had severe psychomotor impairment and needed constant supervision due to frequent seizures. He produced five-word sentences, correctly named numbers up to 30, placed physical numbers in ascending order from 1 to 20 (but could not sequence in descending order), labeled human facial features (but could not position them correctly), named all alphabet letters, and read two-letter syllables (but not three-letter words). Andrew's family requested support to continue school activities and behavioral management that had been interrupted due to the health emergency lockdown. The study took place in Andrew's home.

Experimental Design and Data Collection
The researchers used a multiple-baseline-acrossbehaviors experimental design. Target skills were selected from an interview conducted with parents. The selected targets were ordering numbers 1-20 in descending sequence, positioning eyes, nose, mouth, and ears on an oval face, and reading words.
Each session was five trials, during which the experimenter counted correct and incorrect responses. For ordering numbers, each trial consisted of selecting the flashcard, with correct responses defined as selecting the number that preceded the number of the flashcard placed on the table (e.g., selecting "5" given "6" on the table). For positioning facial features, each trial consisted of positioning a flashcard representing a part of the face in its correct position. For reading, each trial consisted of reading a first-grade level word printed on a flashcard; each word was a consonant, vowel, consonant, vowel sequence (e.g., casa).

Interobserver Agreement (IOA)
Two observers independently viewed and scored data for 50% of the sessions. An agreement was scored when both observers documented the same outcome for each discrete trial. The IOA was calculated on a trial-by-trial basis by dividing the number of agreements by the number of trials and multiplying by 100. The IOA was 94% and 100% for the baseline and VSM phases, respectively.

Treatment Fidelity
An observer coded accuracy of five DTT components in 40% of the trials conducted during baseline and VSM phases (see Supplemental Materials); obtained fidelity was 100%. IOA on fidelity was calculated for the 50% of all baseline and VSM sessions; IOA coefficients were 96% and 99% for the baseline and VSM phases, respectively.

Materials
The experimenter used online meeting software (Skype) via a laptop with an integrated video camera. The participant's mother connected via a desktop computer with an external video camera. The experimenter exported the recordings to an external hard drive.
Instructional stimuli were printed on 6 x 9 cm flashcards. Stimuli in the video model differed from those during the intervention (see Supplemental Materials for details of the stimuli in the video). The videos were recorded at the end of the baseline phase and were created by recording work sessions during which the mother presented prompted each of the three target skills. Each video lasted about 1 min and showed the mother and Andrew sitting at the table next to each other. During the video-recording session, Andrew's mother presented the verbal request (e.g., "Where does the nose go?"), waited 3 s and then prompted a correct answer., which resulted in praise. Incorrect answers or lack of a response in 3 s resulted in physical guidance to emit the correct response. All model and physical prompts were edited from the videos to create the final video model; each video showed the request, the correct behavior, and verbal praise provided by Andrew's mother. Details about the videos appear in the Supplemental Materials.

Interview
During the first meeting, the researchers conducted a structured interview with Andrew's parents based on the Vineland Adaptive Behavior Scale (Sparrow et al., 2016). The interview lasted approximately 2 hr and consisted of 45 open-ended items in three assessment areas: preferences, risk, and academic skills. The parents' answers informed the video modeling strategy and the target skills selected. Parents reported preferences for watching cartoons, movies, and videos depicting Andrew with his parents and relatives. They also reported that Andrew tended to avoid tasks by moving away from the workstation, and that he had difficulties with mathematical tasks, reading, and spatial placement of stimuli.

General Procedure
During the baseline and VSM phases, Andrew's mother presented requests for the tasks measured using discrete independent trials; first, she got her son's attention, then presented the request. If Andrew answered correctly, she provided praise. If Andrew responded incorrectly, she did not provide a correction and moved on to the next trial.

Baseline
For the number task, Andrew's mother presented flashcards with various number sequences (e.g., 15-16-17-18-19-20 or 11-12-13-14-15-16) and instructed Andrew to order them in decreasing sequence. For the facial-feature placement task, the mother presented photographs of each of the four body parts and instructed him to place them on the correct location. For the reading task, the mother presented different words than those shown in the videos, placing them on the table. The experimenter used a collection of 17 words; the same stimuli used in the baseline phase were used in the VSM phase.

Role-Playing and Performance Feedback
Before starting with VSM, Andrew's mother was instructed to use VSM through a few online roleplaying and performance feedback sessions. The experimenter modeled the mother's role with the mother roleplaying the participant's role, then they exchanged. At the end of each role-playing session, the experimenter provided feedback. Teaching continued until the mother performed 90% of steps correctly.

VSM
The intervention phase began 1 week after the end of the baseline phases for each target skill. During VSM, his mother presented Andrew with a video for each session, making sure to alternate the videos between sessions. At the end of the video, she continued with other activities, such as coloring, to avoid potential echoic responses. After 10 min, Andrew's mother evaluated his performance by presented the task using the same materials and procedures used during the baseline phase. Figure 1 shows the results. The data are plotted as the number of correct responses in each session (of five opportunities to respond). Andrew engaged in zero correct responses across all baseline trials for all three target skills. There was an immediate increase in the level of accuracy for each skill during the first session of intervention for each skill. All data points in the intervention phase were higher than those of the baseline phase, and there were no overlapping data points between phases.

Results and Discussion
We propose that VSM may be a replicable intervention model for professionals working via telehealth. Moreover, during the training phase with VSM, his parents reported that they had observed higher compliance levels than the case during the tasks' performance before implementing the intervention. They also reported that they participated with pleasure in the study. The feedback from Andrew's parents, albeit anecdotal, seem to support the social validity of the intervention. Andrew also sat down and watched with interest the videos proposed to him, suggesting he may have found the intervention socially valid. Future research could evaluate participant preference for VSM and systematically evaluate potential collateral effects of the intervention, such as increases in compliance, decreases in challenging behavior, and indicators of positive affect.
Previous research suggests that VSM improves social communication skills, functional skills, academic skills, and behavioral functioning for students with disabilities, especially those with autism spectrum disorder (Bellini & Akullian, 2007;Prater et al., 2012;Boon et al., 2020;Marcotte et al., 2020). This study expands research on the use of VSM to teach academic skills for a new population, adolescents with DS. Future research could assess the generality of using this strategy with other students with DS.
Telehealth interventions may be economically sustainable for families even after in-person services resume after the health emergency. The greater degree of involvement of families when using telehealth could also have positive feedback on personal well-being. Observing children's progress firsthand can give them more positive perceptions about their children and themselves, perhaps increasing their level of self-esteem (Halstead et al., 2018). Order numbers in sequence.

Sessions Correct Responses
This study has some limitations. Creating the video required that Andrew engage in the targeted academic skills and that his mother use prompting strategies. It is possible that this part of the study contributed to his mastery of the skills. Also, Andrew had previous experience with stimuli like those used in this study. Future research should isolate the effects of VSM by using novel stimuli. and minimize the potential for skill acquisition when creating the video models. We attempted to limit the impact of these confounds by using different stimuli during the video than experimental sessions (e.g., numbers 1-10 during the video and 11-20 during the experimental sessions).
A further limitation was that we only obtained anecdotal social validity data. Future studies could use formal questionnaires or evaluation scales to assess the degree of social validity. Despite the limitations, VSM implemented by a parent seems like a promising strategy for individuals with DS.

Declarations
Ethical Approval All procedures performed in studies involving human participants were following the institutional and/or national research committee's ethical standards and the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent
Informed consent was obtained from all individual participants in the study.

Conflict of Interest
No authors have a conflict.