INDIVIDUAL TOPIC SEARCH STRATEGY (ITSS) GUIDELINES

ORIGINAL ARTICLE

The Effectiveness of Video Self-Modeling in Teaching
Active Video Game Skills to Children with Autism
Spectrum Disorder

Erkan Kurnaz1 & Mehmet Yanardag1

Published online: 24 March 2018
# Springer Science+ Media, LLC, part of Springer Nature 2018

Abstract Video self-modeling (VSM) is a teaching method in video-based ap-
proaches for children with Autism Spectrum Disorder (ASD), who have a limited
repertoire of leisure skills and tend towards sedentary behaviours. The aim of this
study was to investigate the effectiveness of the video self-modeling procedure in
teaching active video game skill to children with ASD. The study included 4
children with ASD aged 7 years, who participated in the study and were taught
an active video game skill, consisting of multi-step skills (25), in a one-to-one
training format in five sessions per week. A multiple probe design with probe
conditions across subjects was used to analyse the effects of the VSM. The results
of this study showed that VSM was effective in teaching active video game skills
to children with ASD. The playing of the active video game was continued after
the training process during maintenance and generalization probe sessions. In
addition, the social validity data reflected positive results about acceptability of
intervention, appropriateness of the goals, and importance of the outcomes. VSM
could be utilized to teach motor imitation skills and increase the repertoire of
leisure skills, and active video games are recommended to increase level of
physical activity instead of non-active video games for children with ASD.

Keywords Autism . Video self-modeling . Active video game . Leisure skill . Physical
activity

Autism spectrum disorder (ASD) is characterized by persistent deficits in social communi-
cation and social interaction across multiple context, and restricted, repetitive patterns of

J Dev Phys Disabil (2018) 30:455–469
https://doi.org/10.1007/s10882-018-9596-y

* Mehmet Yanardag
[email protected]

1 Research Institute for Individuals with Disability, Anadolu University, 26470 Eskisehir, Turkey

http://orcid.org/0000-0001-6058-1458

http://crossmark.crossref.org/dialog/?doi=10.1007/s10882-018-9596-y&domain=pdf

mailto:[email protected]

behaviour (DSM-5, American Psychiatric Association 2013). Individuals with ASD have
difficulties in maintaining eye contact, sharing their interest and attention with each other,
turn-taking during conversation, initiating-maintaining and terminating peer interaction, and
using play skills to play games of pretend (Cardon 2015). Individuals with ASD had less
physical activity level and lower motivates toward physical education than their peer (Pan
et al. 2011; Savage et al. 2018), and the limited participation in physical activity was related
to poor motor function (Baranek 2002) and needing extra prompts to initiate activity (Reid
et al. 1991). Therefore, these characteristics of individuals with ASD could predispose them
to engage in less physical activity (Reid 2005) and physical based leisure activities (Turygin
and Matson 2014).

Leisure skills and activities have been taught to children with ASD so that they can
enjoy free time. Leisure skills comprise the ability to identify, access, plan, and
participate in activities, and the skills are important due to the positive effects on
distress, coping skills, social involvement, social support, promoting appropriate re-
sponses and psychological well-being (Chan et al. 2013; Garcia-Villamisar and Dattillo
2010). Leisure skills and activities include a solitary hobby (e.g. producing artwork),
pleasurable activities (e.g. visiting tourist destinations), and attending concerts, exhibi-
tions, movies, plays or playing video games (Turygin and Matson 2014; Blum-Dimaya
et al. 2010). In recent years, active video games, or exergaming, has been utilized as a
leisure activity for children and adolescents (Graf et al. 2009). It has more advantages
than traditional video games that are played in a sitting position, in terms of preventing
sedentary behaviours and providing opportunities for the children to increase their level
of physical activity (Baranowski et al. 2008; Daley 2009; Foley and Maddison 2010).
The video modeling has been utilized to teach novel leisure skills to children with ASD
(Blum-Dimaya et al. 2010; Lee et al. 2017; Yanardag et al. 2013).

Video modeling involves the watching of video clips which provide a model for the
child to imitate (Nikopoulos and Keenan 2006). The video modeling has been shown to
be an evidence-based teaching strategy for children with ASD by the National Autism
Center (NAC 2015). One procedure of the video modeling technology is video self-
modeling (Buggey et al. 1999). Video self-modeling (VSM) is an intervention technique
used to alter the frequency, quality and duration of the target behaviour or skill (Buggey
2009). A video is recorded the observer’s spontaneous behaviour over time in different
contexts then editing the video to involve only the target behaviour or skills (Cihak and
Schrader 2008; Collier-Meek et al. 2012). The concept of modeling was first introduced
as a element of the social learning theory, which express that indivduals learn primarily
through observing the behaviour or skill of those around them (Bandura 1997). VSM can
be applied as positive self-review and feedforward to increase fluency of a skill or the
acquisition of a new skill (Buggey 2005; Dowrick 2011). The positive self-review covers
Bbetter^ form or episode among typical behaviours or skills in a task, and it has good
results in the enhancement of the recently achieved skills (Dowrick 2011). Another
procedure of the VSM is feedforward that an image of success is designed to show
achievement beyond the individual’s current capability. The procedure facilitates remark-
ably changes of target behaviour or skill and development of performance (Dowrick
2011). The effectiveness of VSM has been demonstrated across a variety of social
(Buggey et al. 2016; Kabashi and Kaczmarek 2017; Liu et al. 2015), academic (Burton
et al. 2013), vocational (Cihak and Schrader 2008), behaviour management (Nikopoulos
and Panagiotopoulou 2015), and conversation skills (Sherer et al. 2001) for children with

456 J Dev Phys Disabil (2018) 30:455–469

ASD. However, there have been no studies exploring the effect of VSM on teaching
active video games as leisure skills for children with ASD.

The diagnostic features of ASD generally may affect the repertoire of leisure skills and
activities, and these children may have difficulty learning a novel skill through observation
or imitation of their friends or other individuals. Therefore, children with ASD may need
assistance or receive prompt for exploring and learning novel leisure options (Coyne et al.
2011). The purpose of this study was to answer the following research questions: 1. Will the
use of VSM procedure be effective in teaching active video game skills to four children with
ASD? 2. Will these participants display the skills during the maintenance and generalization
phases? 3. Will the intervention study has high social validity?

Methods

Participants

The study involved 3 boys and 1 girl with ASD, aged 7 years. Before the study, written
informed parental consent and verbal approval were obtained from all the parents of the
participants in compliance with the Declaration of Helsinki. The names of the partic-
ipants have been substituted with pseudonyms for the study. Approval for the study was
granted by the Ethics Committee of the university with research proposal number
10819–2015. To be able to teach the active video game skills using the VSM proce-
dure, some prerequisite skills were identified before the intervention: (a) ability to
follow visual and verbal prompts for at least 5 min, (b) ability to watch a video on the
television (TV) or tablet for at least 3 min, (c) ability to imitate motor skills, and (d),
watching images on the screen. All the participants met these criteria.

Sena was a 7-year old girl with ASD. She had been included in an inclusion program
for primary school, and was also receiving both individual special education (two
sessions a week) and group-based special education. She had difficulty in social
integration and communication skills. She had learned the basic concepts (colour,
shape, etc.) and numbers. Her parents reported that she preferred watching TV or
playing games on a tablet when she had leisure time, and her weekly routine did not
involve a regular physical education program.

Ata was a 7-year old boy with ASD. He had been included in a special education
class in a primary school, and was also receiving both individual special education (two
sessions a week) and group-based special education. He had difficulty in social
integration and communication skills. He had learned the basic concepts (colour, shape,
etc.) and was able to count 1–30. His parents reported that he preferred playing games
on his tablet and was not involved in a regular physical education program.

Uğur was a 7-year old boy with ASD. He was receiving both individual special
education (two sessions a week) and group-based special education (five sessions a
week) in a developmental support centre. He could count 1–10 and had learned the
concepts of colour and shapes. His leisure time was only composed of watching TV. He
had difficulty in social integration and communication skills.

Ersen was a 7-year old boy with ASD. He had been included in a special education
class in a primary school, and was also receiving both individual special education (two
sessions a week) and group-based special education. He had difficulty in social

J Dev Phys Disabil (2018) 30:455–469 457

integration and communication skills. He could reply to simple questions and show
numbers (1–10) when it was demanded. His parents reported that he had not any
physical activity or program in his leisure time. None of the participants had any
experience of the VSM procedure and they had not any experience or knowledge about
active video game skills.

Trainers

The first author conducted the video preparation and intervention phases of the research
process. The second author carried out other tasks such as data collection, recording,
analysis, and preparing the manuscript. Both authors have experience of video model-
ing, single subject design, teaching, active video games, and children with ASD.

Setting

All instructional, probes, maintenance and generalization sessions were conducted in a
research laboratory of the research institute for individuals with disabilities. All sessions
were carried out in a one-to-one format, as five sessions a week in the laboratory, which
was 14 m long, 10 m wide, and covered in an acoustic material.

Materials

A digital camera (Canon 7D Mark II DSLR) and video camera (Sony HandyCam)
were used to record the target behaviours of each participant, and a laptop was
used to upload and edit the video images. Movie maker software (Microsoft) and
other softwares (Adobe Premiere CC, Sony Vegas Pro 11, Adobe Audition CC)
were used to process the images. For the playing of the active video game by the
participants, a games console (Microsoft XBOX 360), game (Microsoft XBOX
Adventures) and Led TV (Samsung) were used in the study. Finally, a flash
memory stick, data collections forms, a writing pad, and pencil were used to
collect data.

Dependent Variable

The main purpose of the study was to determine the effects of VSM on the acquisition
of active video game skills, called BMicrosoft XBOX Adventures^ in children with
ASD. Thus, the target skill was described in detail (Table 1) to be able to reach a
consensus on meeting the criteria and interobserver agreement. The applicability of the
task analyses of the target skill was tested before the intervention phase by four
individuals of a different age, gender, and education profile.

The dependent measure was the percentage of steps completed correctly on task
analysis for the Microsoft XBOX Adventures. A correct step was scored when the
participants independently completed the step as defined in Table 1. An incorrect step
was scored when the participants did not perform the step as defined or did not respond
within five second. Percentage of steps completed correctly was determined as the
number of correct steps was divided by the total number of steps in the task analysis
and multiplied by 100% for each participant.

458 J Dev Phys Disabil (2018) 30:455–469

Experimental Design and Procedure

To determine the effects of VSM on the acquisition of active video game skills in
children with ASD, a multiple probe design with probe conditions across subjects was
used in the study (Kazdin 2011). The research process involved the following steps: the
first baseline, the second baseline (after making video sessions), intervention
(instruction) sessions, maintenance and generalization sessions (Goh and Bambara
2013). The dependent variable of the research was defined as the percentage of the
correctly performed steps of the active video game skill.

Probe Sessions

Baseline probe sessions were conducted as the first baseline and second baseline. The
purpose of the first baseline probe was to determine the independence level of the

Table 1 Task analyses for active video game skill

1. Participant takes the TV remote control from the table.

2. Participant pushes the red button of the remote control to switch the TVon.

3. Participant replaces the remote control on the table.

4. Participant pushes the button on the game console to switch the console on.

5. Participant pushes the button on the CD driver of the console to switch the driver on.

6. Participant turns on the CD box of the game.

7. Participant takes the game CD from the box.

8. Participant places the CD in the CD driver with the picture side uppermost.

9. Participant pushes the button on the CD driver of the console to switch the driver off.

10. Participant walks to the play area marked by yellow tape on the ground.

11. Participant follows some directions on the TV screen for one minute.

12. Participant waits five seconds to adjust the player position.

13. Participant elevates (300) the left arm to the side and waits five seconds when a line character shows on the
TV screen.

14. Participant brings the hand symbol to the start symbol on the screen using his/her hands and waits for five
seconds.

15. Participant follows the task directions on the screen for one minute.

16. Participant initiates active video game by jumping when the video character is seen on the screen.

17. Participant plays the active video game by imitating the video character’s movements (jump, tilt to the
right and tilt to the left)

18. Participant elevates (300) the left arm to the side and waits five seconds to exit from the active video game.

19. Participant pushes the button on the CD driver of the console to switch the driver on.

20. Participant takes the game CD from the box.

21. Participant puts the game CD into the box.

22. Participant pushes the button on the game console to switch the console off.

23. Participant takes the TV remote control from the table.

24. Participant pushes the red button of the remote control to switch the TVoff.

25. Participant replaces the remote control on the table.

J Dev Phys Disabil (2018) 30:455–469 459

participants for the target skill; three baseline sessions were conducted on three
consecutive days. The purpose of the second baseline probe was to determine any
incidental learning during making the video clips (Goh and Bambara 2013). A single
opportunity procedure was used during the baseline and daily probe sessions. The
procedures for the baseline sessions were (a) providing a verbal cue to gain the
participant’s attention, (b) the participant was then asked to perform the skill, (c) five
seconds were allowed for the participant to initiate the skill, (d) If the participant
displayed the correct behaviour, the performance was recorded as plus (+) for that
target skill, (e) If the participant performed an incorrect response, the performance was
recorded as a minus (−), and the assessment was terminated (Brown and Snell 2000).
After the probe procedure, praise was delivered at the end of the probe regardless of the
participant’s performance.

Similarly, daily and full probe sessions were conducted exactly as the baseline
sessions. The daily probe session was performed to determine the current per-
formance of the participant to decide whether to stop the intervention process.
The daily probe session was useful to observe when the criterion of the target
skill had been reached. The daily probe sessions were regulated as one trial after
each intervention session, and were maintained until a participant showed 100%
correct performance in three consecutive daily probe sessions for the target skills.
Full probe sessions were conducted before the intervention, and after reaching
the criterion of the target skill (it was determined by the daily probe sessions),
and were continued until a stable response had been demonstrated in at least
three consecutive sessions. The full probe sessions were terminated after
collecting the full probe data from the last participant.

Making Video Sessions

VSM clips were prepared after the first baseline sessions. For this purpose, video
images were recorded while demonstrating the steps of the skill to all the
participants. The recording speed of the videos was 18–135 mm. A Canon 7D
Mark II DSLR camera and Sony HandyCam video camera and tripod were used
in these sessions. Researchers, participants and ancillary staff were involved in
these sessions. The participants were shown the steps of the target skill by the
ancillary staff. A partial physical prompt was delivered by the staff to the
participants to display some steps of the skill during the video recordings. To
prevent possible an incidental learning, complex step sequences and different
video recording angles were used during the sessions. Each step of the skill was
recorded by two video cameras placed at different angles, and then these
recordings were converted to form the VSM clip, which was used in the
intervention phase of the research. The VSM clip was evaluated for validity by
two special education experts. As a result of the expert evaluation, the VSM clip
was specified as valid and understandable for the intervention process. The
average duration of the video clip recording sessions for each participant was
calculated as 15 h. It has been stated in literature that the length of the VSM clip
in intervention sessions should not be longer than 5 min (Nikopoulos and
Keenan 2006). In this study, the average length of the VSM clip was 3 min
25 s (range: 3 min 10 s – 3 min 45 s).

460 J Dev Phys Disabil (2018) 30:455–469

Intervention Sessions

The VSM process for the active video game skill was initiated when the steady-state
data was obtained after the second baseline probe session. The intervention sessions
were performed as follows: (a) The researcher delivered an attentional cue (Bare you
ready to watch a video clip?^) to the participant to focus on the task. The researcher
provided verbal praise (Bwonderful^) if the participant replied verbally or with a sign.
(b) The researcher and participant sat side by side at a table, and the participant watched
the self-modeling video clip on a tablet. After watching the clip, the researcher asked,
Bthe video has finished, would you like to watch it again?^ and waited for five seconds.
If the participant wanted to watch the video clip again, the researcher provided the
opportunity of watching it twice more. While watching the video clip, researcher did
not give any information about the images to the participant. When the participant
turned away from the screen, the researcher delivered a verbal prompt (Bwatch the
video^) to maintain attention on watching the video clip. When the participant watched
the video clip, the researcher provided verbal praise (Bwell done^). (c) The researcher
and participant moved to the corner of the laboratory where the game console and TV
were placed in order to perform the active video game immediately after watching the
VSM clip. (d) The researcher provided an attentional cue (Bare you ready to play the
video game?^) to the participant to focus on the task. (e). The researcher provided task
direction (Bplay^) if the participant replied verbally or with a sign. The researcher
delivered verbal reinforcement (Bwell done, bravo^) when the participant performed
correctly the steps of the active video game skill. An intervention session was termi-
nated if a participant performed an incorrect response or gave no response for five
seconds. Two trials were performed for the target skill in every intervention session,
and a one-hour break was given between the two trials.

Maintenance and Generalization Sessions

Maintenance sessions were implemented one, two, and four weeks after the participants
met the instructional criterion, which was to display independent performance for the
target skill. Maintenance sessions were performed in a similar manner to the probe
sessions. Generalization across settings was evaluated by pre-post-test design. This was
applied as a probe session before the intervention process and after the participants had
reached the criteria for the active video game skill.

Data Collection and Analysis

Four types of data were collected during the study including effectiveness, mainte-
nance, generalization, social validity and reliability data. Effectiveness, maintenance,
and generalization data were collected via recording data of children’ correct and
incorrect responses for the target skill and calculated the percentage of correct re-
sponses. The collected data of effectiveness, maintenance and generalization phases
was analysed graphically.

The reliability data was collected as interobserver agreement and procedural reli-
ability from 30% of all the experimental sessions. Interobserver agreement was calcu-
lated using the point-by-point method with a formula of the number of agreements

J Dev Phys Disabil (2018) 30:455–469 461

divided by the number of agreements plus disagreements multiplied by 100 (Kazdin
2011). Interobserver agreement data were collected in full probe, daily probe, inter-
vention, maintenance and generalization sessions for each participant. Procedural
reliability (independent variable) was calculated by dividing the number of observed
researcher behaviours by the number of planned researcher behaviours multiplied by
100 (Kazdin 2011). The researcher’s behaviours were observed as follows: (1) record-
ing the video clip, (2) providing the attentional cue before watching the video clip, (3)
delivering the self-modeling video images, (4) providing prompting/reinforcement
while watching the video, (5) preparing materials, (6) providing the attentional cue
before delivering task direction, (7) providing task direction, (8) delivering reinforce-
ment after performing the task, (9) delivering verbal praise for collaboration.

Social validity data was collected to determine the parents’ opinions about study. In
order to gather data from parents, social validity questionnaire form was developed.
The form consisted of 6 closed questions (1. Do you think that the active video game
skills which was taught in this research is important for your child? 2. Do you think that
the active video game skill which was taught in this research will contribute to
repertoire of the leisure skill of your child? 3. Do you satisfy to participate in the study
that was taught active video game skills to your child by using video self modeling? 4.
Do you think that your child learned the leisure skill? 5. Do you intend to buy an active
video game console so that your child can maintain active video game skill during his/
her leisure time in home? 6. Would you consider re-joining a new intervention study
that aims to teach a novel skill to your child by using VSM?) and two open ended
questions (1. Would you describe the parts of the study that you approve? 2. Would you
describe the parts of the study that you disapprove?). Parents were asked to write the
answer of the questions under the items. The social validity forms were provided in
closed envelopes to the parents and they were asked to reply and send the answers back
in closed envelopes to the researchers. Answers of these questions were analysed
descriptively.

Results

Interobserver Agreement and Procedural Reliability

Interobserver agreement data revealed that both researchers agreed on the participants’
performance during the first baseline probe session, second baseline probe session,
daily probe session, full probe session, maintenance and generalization sessions, and it
was calculated as 100% for all participants. Procedural reliability was calculated, and
the researcher performed the above-mentioned procedures at 92% (range: 88–100%)
for all participants.

The Effects of the Video Self-Modeling on the Active Video Game Skill

Figure 1 shows the effectiveness data as performance of the participants in the active
video game skill during baseline, intervention, maintenance, and generalization

462 J Dev Phys Disabil (2018) 30:455–469

FB MVS SB I FBS1 I FBS2 I FBS3 I FBS4 M
P

er
ce

nt
o

f c
or

re
ct

r
es

po
ns

es

Sessions

Fig. 1 Percent of correct responses in video play skill for all participants. FB: First baseline SB: Second
baseline MVS: Making video sessions I: Intervention sessions FPS: Full probe session M: Maintenance

J Dev Phys Disabil (2018) 30:455–469 463

sessions. Each data point shows the percentage of correct responses during the sessions.
As seen in Fig. 1, Sena displayed 0% performance for the active video game skill in the
first and second baseline probe sessions. She showed an increase in the percentage of
independent skill when VSM was provided, and she reached the acquisition criterion of
100% in the third intervention session. When she showed 100% performance in three
consecutive sessions, the intervention sessions were terminated. The researcher con-
ducted five sessions and ten trials in the intervention phase. She maintained the
performance (100%) during all full probe and follow-up sessions. She also displayed
100% performance for the active video game skill when the generalization sessions
were evaluated.

As seen in Fig. 1, Ata displayed 0% performance for the active video game skill in
the first and second baseline probe sessions. He showed an increase in the percentage of
independent skill when VSM was provided, and he reached the acquisition criterion of
100% in the fourth intervention session. When he showed 100% performance in three
consecutive sessions, the intervention sessions were terminated. The researcher con-
ducted six sessions and twelve trials in the intervention phase. He maintained the
performance (100%) during all full probe and follow-up sessions. He also displayed
100% performance for the active video game skill when the generalization sessions
were evaluated.

As seen in Fig. 1, Ugur displayed 16% performance for the active video game skill
in the first and second baseline probe sessions. He showed an increase in the percentage
of independent skill when VSM was provided, and he reached the acquisition criterion
of 100% in the sixth intervention session. When he showed 100% performance in three
consecutive sessions, the intervention sessions were terminated. The researcher con-
ducted eight sessions and sixteen trials in the intervention phase. He maintained the
performance (100%) during all full probe and follow-up sessions. He also displayed
100% performance for the active video game skill when the generalization sessions
were evaluated.

As seen in Fig. 1, Ersen displayed 0% performance for the active video game skill in
the first and second baseline probe sessions. He showed an increase in the percentage of
independent skill when VSM was provided, and he reached the acquisition criterion of
100% in the sixth intervention session. When he showed 100% performance in three
consecutive sessions, the intervention sessions were terminated. The researcher con-
ducted six sessions and twelve trials in the intervention phase. He maintained the
performance (100%) during all full probe and follow-up sessions. He also displayed
100% performance for the active video game skill when the generalization sessions
were evaluated.

Social Validity Results

Social validity questionnaire forms were obtained from parents after the intervention
process. Six closed and two open ended questions were asked to parents in order to
determine the social validity data of the study. In the first closed question, all parents
stated that playing active video game skills was important for their child. In the second
closed question, all parents mentioned that the skill will contribute their repertoire of
the leisure skills because they need some alternative skills during at home to occupy
actively himself/herself. In the third closed question, all parents stated that they pleased

464 J Dev Phys Disabil (2018) 30:455–469

to engage in this study owing to the target skills to create opportunities to play with
peers. In the fourth closed question, all parents affirmed that all children learned the
new leisure skill. In the fifth closed question, three parents stated that active video game
console is needed because of maintaining the skill in the home and would buy it,
whereas the forth parent told that it was not suitable for their economic condition. In the
sixth …