The Effectiveness of a Packaged Intervention Including Point-of-View Video Modeling in Teaching Social Initiation Skills to Children With Autism Spectrum Disorders

Abstract

Deficits in social communication and interaction have been identified as distinguishing impairments for individuals with an autism spectrum disorder (ASD). As a pivotal skill, the successful development of social communication and interaction in individuals with ASD is a lifelong objective. Point-of-view video modeling (VM) has the potential to address these deficits. By presenting only what a person might see from his or her viewpoint, it may be more effective than other forms of VM, in limiting irrelevant stimuli and providing a clear frame of reference to facilitate imitation. The current study investigated the use of point-of-view VM in teaching social initiations (e.g., greetings). Using a multiple baseline across participants design, five kindergarten participants were taught greetings using a packaged intervention, which included point-of-view VM, video priming, verbal praise, reinforcement, and prompting. Immediately before and after viewing the entire point-of-view video model, the participants were evaluated on their greetings with a trained, typically developing peer serving as a communication partner. Specifically, the greetings involved participants’ abilities to shift their attention toward the peer who entered the room, maintain attention toward the peer, and engage in an appropriate greeting (e.g., hi, hello). Both generalization and maintenance were tested. Overall, the data suggest point-of-view VM is a promising intervention for increasing greetings. However, both generalization and maintenance were limited. Despite the limitations of the study and variable results, there are a number of implications moving forward for both practitioners and future researchers examining point-of-view modeling.

Keywords

autism spectrum disorder point-of-view video modeling first-person perspective video modeling social initiations greetings social communication and interaction

A predominant component of the criteria for autism spectrum disorder (ASD) published in the fifth edition of the Diagnostic and Statistical Manual (5th ed., DSM-5; American Psychiatric Association, 2013) is social communication and interaction skills. This diagnostic criterion addresses three symptoms: (a) deficits in social-emotional reciprocity, (b) deficits in nonverbal communicative behaviors used for social interaction, and (c) deficits in developing and maintaining social relationships (American Psychiatric Association, 2013).

The critical importance of social communication and interaction has impelled research on interventions targeting such skills. However, despite the social skills interventions available for individuals with ASD, more research on interventions is warranted (Jung & Sainato, 2013; White, Keonig, & Scahill, 2007). In a comprehensive review of the intervention research on social development that spanned from 1985 to 2006, White et al. (2007) concluded there was still much to research with regard to effective intervention approaches. The authors also emphasized a need to conduct replication, extending, and more methodologically rigorous studies.

Among the interventions identified as being evidence-based by the National Professional Development Center on Autism Spectrum Disorders (2014) necessitating further research is video modeling (VM). Using recorded videos demonstrating a targeted skill or behavior, VM provides individuals multiple opportunities to observe and imitate the video (Hine & Wolery, 2006). VM may be presented through video priming, which is when the video is presented in its entirety prior to imitating and practicing the desired skill, or video prompting, which involves segmenting the video into steps to scaffold the learning of the skill (Mason, Davis, Boles, & Goodwyn, 2013). As a progressively common instructional tool by educators, due to increased access to technology within the school context, VM addresses the characteristic preference for visual stimuli shown by this population (Shane et al., 2012).

There are four identified forms of VM: (a) adult modeling, (b) peer modeling, (3) video self-modeling, and (d) point-of-view VM. The latter is also known as first-person perspective VM. Point-of-view VM is a video of what an individual would see if he or she were engaging in the targeted behavior or skill. Shukla-Mehta, Miller, and Callahan (2012) have identified point-of-view VM as a promising intervention for individuals with ASD. Considering the challenges of the population with theory of mind (Baron-Cohen, Leslie, & Frith, 1985) and attending to relevant cues, filming a VM from the individual’s perspective may better support the imitation of the targeted skill by limiting irrelevant stimuli and directing the individual’s attention to the relevant stimuli (Barton, Lawrence, & Deurloo, 2012; Rayner, Denholm, & Sigafoos, 2009; Tetreault & Lerman, 2010).

A comprehensive review of the literature conducted by Lee (2015) found the application of point-of-view VM to teach social interaction and communication to be limited, but promising. Hine and Wolery (2006) investigated whether point-of-view VM could be used to teach appropriate play with two toy sets (i.e., gardening and cooking) containing various items. A point-of-view video model, which had an introductory 2-min cartoon clip, showed two hands appropriately playing with each of the toy sets. Looking specifically at whether the participants imitated the actions modeled in the video, the authors found the intervention to be effective. However, one participant required additional prompting and reinforcement to acquire the skills, whereas the second participant was able to master both sets. Generalization from the therapy room to the classroom occurred only with the gardening set, and data from the maintenance probes were variable. Replication studies, further research of necessary participant prerequisite skills, and further investigation into how gained skills may be generalized and maintained are significant implications.

Sancho, Sidener, and Reeve (2010) used the intervention to teach two participants to play appropriately with two play sets (i.e., house and circus). The point-of-view video model presented 10 scripted actions, which involved adult hands manipulating the characters in the set, and 10 vocal scripts. The authors also recorded the video from additional angles. While viewing the videos, participants were prompted to play with the targeted set simultaneously. The intervention was effective for the participants; however, any elaboration on the scripted actions and vocal scripts was minimal. Based on the study, there is a need for replication, further comparison of video priming and prompting, and facilitation of generalization in future research.

Tereshko, MacDonald, and Ahearn (2010) investigated the impact of point-of-view VM on teaching four participants functional play skills with Mega Bloks®. Each video model included adult hands creating a toy structure with the blocks. These videos were then segmented into individual steps, thus creating video prompts. All participants were able to imitate and build the toy structures and also generalize the skills in the classroom. As with the previous studies, further investigation of maintenance is warranted.

Most notably, Tetreault and Lerman (2010) examined how point-of-view VM may teach three children diagnosed with autism to initiate and maintain social communication. The three social scripts, which were modeled using the intervention and were accompanied by related materials, were titled “Get Attention,” “Request Assistance,” and “Share a Toy.” The point-of-view video involved the first author verbalizing the script from behind the camera. An unfamiliar adult was the conversant and also verbalized portions of the script. The video model also moved in a manner to mimic head movements. The novel application of point-of-view VM to teach social interaction in this study led to inconclusive results. Supplementary reinforcement and prompting, and adjustments to the scripts were necessary for two of the three participants to engage in the social exchanges. Generalization of social exchanges with novel materials was minimal, and maintenance was variable. The authors noted that identifying which components of the study were necessary to generate the desired results needed further examination. The authors also discussed how the complexity of the social scripts may be linked to the variable performance and emphasized the need for future research to further apply point-of-view VM to teach social communication.

Mason et al. (2013) indicated this form of VM was promising for individuals with ASD, which aligned with the conclusions of past literature reviews of the efficacy of VM by McCoy and Hermansen (2007) and Shukla-Mehta et al. (2012). Yet, despite these statements, there continues to be limited research on this type of VM and a number of unanswered questions, which include fading of prompting and reinforcement, and how both generalization and maintenance of social skills can be enhanced. In addition, Tetreault and Lerman (2010) surmised that,

Further analysis of the usefulness of the [point-of-view video modeling] technique to teach social skills to children with autism is needed. An intermediary step between simple social skills (e.g., greetings) and more complex skills like those assessed here is warranted. (p. 416)

Based on Tetreault and Lerman’s inconclusive results, the current investigation aimed to take a prerequisite approach and address how effective this form of VM may be in teaching young children with ASD to first initiate a greeting. Specifically, the following questions were addressed in the research:

Research Question 1: To what extent did appropriate greetings increase as a result of point-of-view VM intervention presented with video priming, verbal praise, reinforcement, and prompting?

Research Question 2: To what extent did procedures to fade reinforcement and prompting lead to independent performance of the appropriate greetings?

Research Question 3: To what extent did any increased greetings generalize across novel settings and peers not included in the research study?

Research Question 4: To what extent did any increased greetings maintain 2 and 4 weeks following the conclusion of the study?

Method

Participants With ASD

Five male, kindergarten participants, who met the criteria for ASD according to the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM-IV-TR; American Psychiatric Association, 2000), were selected for the study. Through observations and conferencing with the classroom teachers, the participants selected for the study exhibited all the characteristics and prerequisite skills outlined in Table 1. All participants were observed engaging in limited social–emotional reciprocity, nonverbal communicative behaviors, and developing and maintaining relationships, which included making friends and demonstrating interest in other peers. In addition, Participant 1 frequently engaged in perseverative speech and highly restricted interests (e.g., television shows, videos, animals). Participant 2 was observed frequently engaging in perseverative speech, excessive adherence to routines, and engagement in self-stimulatory behavior (e.g., body rocking, flapping of hands). Both Participants 4 and 5 also engaged in perseverative speech, and Participant 5 engaged in stereotypic behavior (e.g., flapping of hands). Additional assessment scores were collected from students’ administrative records, which are identified in Table 2.

Table 1.

Characteristics and Prerequisite Skills of Participating Students.

1. Exhibits low rates of disruptive behaviors
2. Requires minimal prompting (e.g., gestures, verbal) across a majority of tasks
3. Uses at least two word phrases with adults and peers
4. Responds to “Do you want . . . ?” questions by verbally answering yes or no
5. Answers who and what wh-questions with at least a one word response
6. Has been observed verbally initiating interactions with typically developing peers and verbally respond to peer initiations
7. Engages in emerging or basic imitation skills (e.g., vocalizations, body movements, object use)
8. Technology (e.g., iPad, tablet, computer) has been used as a reinforcer for correct responding
9. Exhibits the ability to attend to a video for approximately 3–4 min
10. Has social communication and interaction goals in their individualized education program

Table 2.

Descriptive Information on Participating Students.

Participant	Age	ASRS	CARS-2	Leiter-3Nonverbal IQ	Receptive language(years)	Expressive language(years)	Social/emotional behavioral(years)
Participant 1	6.2	T: 74 P: 77	—	—	2–3^a	2–3^a	2–3^a
Participant 2	6.6	P: 69	42	—	2–3^a	2–3^a	2–3^a
Participant 3	5.5	T: 76 P: 79	—	72	2–3^a	1–2^a	1–2^a
Participant 4	5.6	T: 78 P: 77	—	93	3–4^a	3–4^a	2–3^a
Participant 5	5.10	T: 78 P: 64	35	—	3–4^b	3–4^b	2–3^b

Note. ASRS = Autism Spectrum Rating Scale (Goldstein & Naglieri, 2012); CARS-2 = Childhood Autism Rating Scale–2 (Schopler & Van Bourgondien, 2010); Leiter-3 = Leiter International Performance Scale–3rd Edition (Roid, Miller, Pomplun, & Koch, 2013); T = teacher rating; P = parent rating; WSS = Working Sampling System.

Early Childhood Skills Development Guide and WSS (Meisels, Marsden, Jablon, Dorfman, & Dichtelmiller, 2001). ^bBDI-2 = Battelle Developmental Inventory–2nd Edition (Newborg, 2004).

Communication Partners

Communication Partner 1 was a 6-year-old, male, kindergarten student, and was filmed as part of the point-of-view VM. Communication Partner 1 also took part in the baseline phase probes, intervention phase probes (i.e., daily probes, practice session probes), post-intervention phase probes, and maintenance phase probes. Communication Partner 2 was a 5-year-old female, kindergarten student, and was selected to participate in the generalization probes. Both students received training from the investigator.

Settings

The study took place in a public elementary school. The participants were students in a self-contained classroom that included 12 students, two special education teachers, four paraprofessionals, and an occasional volunteer. The program used both whole-group and small-group teaching procedures to target academic skills. The students also received whole-group speech instruction for approximately 1 hr each week.

The probes in the baseline phase and postintervention phase were conducted in the participants’ classroom. The generalization probes within the baseline phase and the generalization phase following the intervention were also conducted in the classroom and in the school library, art room, music room, and/or computer lab. The intervention and all probes in the intervention phase (i.e., daily probe, practice session) occurred in two similar intervention rooms in the school where distractions could be minimized. Maintenance phase probes were conducted in the participants’ classrooms and also in the intervention room.

Independent Variable

The independent variable was a packaged intervention that consisted of a point-of-view video, which modeled the greeting, along with the use of verbal praise, reinforcement, and prompting. Filming the point-of-view video involved the use of an iPad camera. The camera was first directed away from the intervention room door. The sound of the room door being opened and then shut occurred and the camera turned in the direction of the door, which mimics turning of the body and attention toward an individual entering the room. In the video frame was Communication Partner 1. From behind the camera, the investigator stated, “Hello,” which was the verbal greeting to be taught to the participants. Then Communication Partner 1 smiled, waved, and responded by saying, “Hello.”

An approximately 10-s visual introduction of a cartoon clip was inserted before the actual point-of-view video. The video clip was of a puppet singing a counting song, and was intended to gain the attention of the participant and prepare the participant to view the point-of-view video model. This approach has been used in previous research (Hine & Wolery, 2006; Tetreault & Lerman, 2010) to help the participant to attend to the video. The total length of the point-of-view video, including the introductory video clip, was 22 s.

Dependent Variable

The dependent variable was the percent of correct responses to three behaviors required to engage in the greeting. The three measured behaviors of the dependent variable included (a) shifting of attention toward communication partner who is entering the room, which was identified as correct if the participant turned to look at the communication partner within 5 s of the door being shut; (b) engaging in the greeting, which was identified as correct if the participant said, “Hello,” or verbalized any variation of a greeting (e.g., hi, hey) within 5 s of the door being shut by the communication partner; and (c) maintaining attention toward the communication partner, which was identified as correct if the participant continued to look in the direction of the communication partner while also verbalizing the greeting.

Experimental Design

A multiple baseline across participants design was used to assess the effectiveness of point-of-view VM in teaching greetings to kindergarten children with ASD. The design exhibits experimental control by demonstrating that the dependent variable increases only when the independent variable is applied. Thus, a firm relationship between point-of-view VM and greetings may be established if greetings increase as the independent variable is applied successively to the target participants’ behavior.

Procedures

Baseline phase

Baseline phase probes were conducted in each participant’s classroom. The participant first stood or sat in a chair positioned away from the classroom door. Communication Partner 1 entered the classroom. After 5 s of Communication Partner 1’s entry, the participant’s engagement in the three measured behaviors of the dependent variable was recorded as correct or incorrect. Two to four baseline phase probes per participant were conducted daily and each probe was conducted in approximately 1 min. The investigator did not interact with the participants during these probes, and no reinforcement or instruction was provided during this phase of the study.

Intervention phase

The order in which participants received intervention was determined randomly, with student availability being a factor. Immediately prior to beginning the first intervention session of that school day, a daily probe was conducted, applying the same methodology as baseline (cf. Hine & Wolery, 2006) but in the intervention room. The purpose of the daily probe was to identify the participant’s performance without immediately seeing the video prior to practicing the targeted behaviors.

Two to four intervention sessions occurred daily, and involved the delivery of the intervention and conducting the practice session probe in the intervention room. Each session began by having the participant watch the entire point-of-view video on an iPad. The investigator provided verbal praise to the participant for attending to the video model. If the investigator provided gestural prompts, had to block the participant from touching the iPad screen, or any combination of these 3 times, the video was stopped and restarted from the beginning. No instructional statements elaborating on the video model were made during the delivery of the intervention.

A practice session probe immediately followed the viewing of the video and began with the investigator stating, “Let’s practice!” The participant stood or sat in a chair positioned away from the door. Communication Partner 1 then entered the room, and the three target behaviors of the dependent variable were scored. Verbal praise was provided if the participant attempted to engage in any one of the three target behaviors. Both verbal praise and a tangible reinforcer were provided if the participant accurately engaged in all three target behaviors.

A correction procedure was implemented if the participant responded incorrectly or did not initiate any targeted action or statement. If the participant did not turn in the direction of Communication Partner 1 after 5 s of the door being shut or did not maintain attention toward Communication Partner 1 while verbalizing the greeting, the investigator provided a gestural prompt (i.e., pointed) toward the communication partner who had entered the room. After an additional 2 s with no response or an incorrect response, the investigator provided another pointing prompt accompanied by the verbal prompt, “Look.”

If the participant did not say, “Hello,” within 5 s of the door being shut, the investigator provided a pointing prompt toward Communication Partner 1 accompanied by a partial verbal prompt by first making an /h/ sound. If the participant did not engage in the greeting or responded incorrectly after an additional 2 s, the investigator provided another pointing prompt and a full verbal prompt, “Hello.” For all the correction procedures, if there continued to be no or an incorrect response after an additional 2 s of the full verbal prompt, the practice session probe was concluded and another intervention session began (cf. Tetreault & Lerman, 2010).

During the practice session probes, if the participant performed the behavior during the correction (i.e., after prompting), the investigator provided verbal praise and continued to implement the next correction procedure for the following behavior if that behavior was not performed after 2 s. Any behavior performed after the correction procedure was recorded as incorrect. After the participant reached a 100% criterion level with the daily probes across three consecutive days, the postintervention phase began on the following day.

Postintervention phase

Mirroring baseline phase procedures, data on the participants’ ability to complete all three measured behaviors of the dependent variable were collected in each participant’s classroom. However, following variable postintervention data for Participant 1, retraining occurred in the form of having the participant view the point-of-view video once in its entirety in the classroom setting or in the hallway (i.e., delivery of intervention with video priming; cf. Tereshko et al., 2010). Immediately following retraining, a postintervention probe was conducted in the classroom. Retraining was repeated until Participant 1 met the established criteria and demonstrated the three measured behaviors of the dependent variable across three consecutive postintervention probes. The retraining procedure also occurred for Participants 3 and 5 after the first three consecutive data points in the postintervention phase were calculated at 33% or below and did not demonstrate an ascending trend.

Generalization with Communication Partner 2 and natural generalization

Two methods of collecting generalization data were conducted to determine whether the target behaviors generalized, and were collected both before and after the intervention. Generalization probes were conducted at the beginning and at the end of the baseline phase to measure the occurrence of any greetings with Communication Partner 2 in a novel setting. In addition, four to five generalization probes were conducted immediately following the postintervention phase.

Activities in the classroom (e.g., indoor recess, snack time) were identified, in which natural opportunities for greetings with other peers in the classroom would be likely to occur. The investigator conducted a 5-min observation and measured the frequency of the targeted and any other social initiations. For the purposes of these probes, greetings were targeted, as well as other social initiations, which included getting attention, organizing, sharing, seeking assistance, engaging in compliments, and demonstrating affection. The operational definitions of these target behaviors were modified from the work done by Odom and Strain (1986) on increasing social initiations of individuals with ASD with peers during play opportunities. Two noncontinuous natural generalization probes were conducted in the baseline phase, and four natural generalization probes were conducted immediately following the postintervention phase.

Maintenance

Baseline phase procedures were used to assess maintenance of the target behaviors, and Communication Partner 1 participated in these probes. Maintenance probes were conducted in the classroom and in the intervention room 2 and 4 weeks following the conclusion of the intervention for each participant.

Social validity

All teachers (N = 2), paraprofessionals (N = 3), and parents (N = 4) of participants with ASD were asked to complete a questionnaire on which they rated and explained the extent to which they observed an increase in greetings with others (e.g., classmates, siblings, parents) by the participants.

Reliability

Interrater reliability

Reliability data were collected for 40% of all video-recorded probes across all phases for each of the participants. The investigator provided training for the reliability observer through discussions of the operational definitions of the targeted behaviors of the dependent variable and training videos. The training was concluded once the observer and investigator reached 80% agreement with the training videos.

The reliability observer viewed video-recorded probes and completed an interrater reliability form, which mirrored the data recording form utilized by the investigator. Reliability was calculated by dividing the smaller number recorded by the larger number recorded and multiplying by 100%. The average calculated interrater reliability was 99.3% (range = 0%–100%). It is important to note that there was only one occasion where there was 0% agreement between the investigator and the observer. The disagreement was with regard to a natural generalization probe for Participant 2, where there was disagreement on whether the participant was seeking attention from another peer by stating the peer’s name.

Fidelity of implementation

An independent observer conducted procedural reliability of the video-recorded training sessions for the communication partners. Fidelity observations were also conducted on 40% of video-recorded intervention sessions by an independent observer to ensure that all intervention procedures implemented were conducted with integrity. Reliability was calculated by totaling the number of steps completed, and dividing by the total number of steps necessary, and multiplying it by 100%. The average calculated procedural reliability of the training sessions for the communication partners was 100%, and 96.7% (range = 93.1%–100%) for the intervention across all participants.

Results

The results of the study are represented in Figure 1. With a visual analysis of the baseline phase, none of the participants engaged in any of the targeted behaviors prior to the intervention. For Participant 1, the level did not change, from the baseline phase to the first daily probe in the intervention phase, thus demonstrating a slight overlap between phases. However, from the baseline phase to the first practice session probe (i.e., following the initial introduction of the intervention) in the intervention phase, the level ascended. When examining the last three data points in the baseline phase and the first three daily probes in the intervention phase, immediacy of the effect was observed. Following the first daily probe and the introduction of the intervention, subsequent practice session probes and daily probes immediately ascended, and the data stabilized and remained high at 100%. From the intervention phase to the postintervention phase, there was a descending level change, and the trend in the postintervention phase was highly variable. However, following a single retraining for Participant 1, the trend and level of the postintervention data ascended from 0 % to 100% and stabilized at 100% for three consecutive sessions.

Figure 1.

Multiple baseline design across participants, with percentage of accurately performed behaviors of the dependent variable during baseline, intervention, postintervention, generalization, and maintenance probes.

From the baseline phase to the first daily probe in the intervention phase, the level did not change for Participant 2; however, from the baseline phase to the first practice session probe, the level ascended by 33%. Immediacy of the effect was observed in the baseline phase probes and the daily probes. Following the first 2 days of intervention sessions, the practice session probes were highly variable with repeated ascension and declension. Following the third daily probe, subsequent practice session probes and daily probes ascended and the data stabilized and remained high at 100%. From the intervention phase to the postintervention phase, there was no level change due to the data in the postintervention phase remaining at 100% across four consecutive probes.

From the baseline phase to the first daily probe in the intervention phase for Participant 3, the level did not change, thus demonstrating a slight overlap. However, from the baseline phase to the first practice session probe, the level ascended by 33%, and immediacy of the effect was observed. Following the first daily probe and the introduction of the intervention, subsequent practice session probes and daily probes immediately ascended and remained high at 100%. From the intervention phase to the postintervention phase, there was no level change. However, following that initial probe, the trend in the postintervention phase descended to 33% for three data points. Throughout multiple implementations of retraining, the trend was highly variable and never ascended and stabilized at 100% for three consecutive data points.

From the baseline phase to the first daily probe, the level did not change for Participant 4; however, from the baseline phase to the first practice session probe, the level ascended by 33%. When examining the last three data points in the baseline phase and the first three daily probes in the intervention phase, immediacy of the effect was observed. Following the first daily probe, responding during subsequent practice session probes and daily probes immediately ascended and remained high at 100%. The data in the postintervention phase remained at 100% across four consecutive probes.

From the baseline phase to the first daily probe and the first practice session probe, the level did not change for Participant 5. When examining the last three data points in the baseline phase and the first three daily probes in the intervention phase, immediacy of the effect was observed. The data in the first 2 days of intervention sessions were highly variable, with an ascension from 0% to 33% and then a return to 0%. Then, following the second daily probe, which was scored at 100%, there was a sudden declension to 0% and then immediate ascension to 100%. The data then stabilized and remained at 100%. From the intervention phase to the postintervention phase, there was a descending level change, and the trend in the postintervention phase was highly variable throughout multiple implementations of retraining. However, following the third retraining for Participant 5, the trend of the postintervention data ascended and remained high at 100% for three consecutive data points.

Effect Size Calculations

The calculated effect sizes for percentage of data points exceeding the median (PEM) for participants is displayed in Table 3. PEM scores for all five participants fell within the fairly effective treatment and moderately effective treatment range, respectively. The calculation for percentage of all non-overlapping data (PAND) = 96.1% and the translated phi coefficient (φ) = 0.87 suggests a strong positive relationship. However, it is important to note the limitation of these effect sizes, especially because the calculation of effect sizes within single-subject research is in its infancy and remains controversial. Horner, Swaminathan, Sugai, and Smolkowski (2012) stated the following:

to date, however, no statistical approach for examining single-case research has met three fundamental criteria: (a) controls for auto-correlation (e.g. the fact that scores are not independent), (b) provides a metric that integrates the full constellation of variables used in visual analysis of single-case designs to assess the level of experimental control demonstrated by the data, and (c) produces an effect-size measure for the whole study (as opposed to two adjacent phases). (p. 270)

Despite the limitations and need for further discussion of the application of statistical analyses in single-case research, effects sizes were calculated to further support the visual analysis of the graphed data. In addition, the inclusion of calculated PEM and PAND may facilitate future meta-analyses involving single-case research.

Table 3.

Calculated Effect Sizes.

Participant	PEM
Participant 1	0.75
Participant 2	0.83
Participant 3	0.80
Participant 4	0.80
Participant 5	0.75

Note. PEM = percentage of data exceeding the minimum.

Generalization

Generalization with Communication Partner 2

For all participants, performance during the generalization probes with Communication Partner 2 was 0% in the baseline phase. Participant 1’s average performance during the generalization probes following the intervention was 6.6% (range = 0%–33%). Participant 2’s and Participant 3’s performance remained at 0% after the intervention. Participant 4’s performance in the generalization probes following the intervention was 33%, and Participant 5’s performance was 24.8% (range = 0%–33%).

Natural generalization

For all participants, there was no performance of the targeted greeting or other social initiations for the natural generalization probes conducted in baseline. For Participant 1, the average performance for the natural generalization probes in the postintervention phase was 0.75. Participant 1 was observed requesting for toy cars from his peers, which was identified as sharing. For Participant 2, the average performance was 0.50. Participant 2 was observed seeking the attention of one female peer by saying her name, which was identified as attention. For Participant 3, neither the targeted greetings nor the other social initiations occurred in the postintervention phase. The average performance for Participant 4 in the postintervention phase was 0.50. Participant 4 was observed greeting other peers, which was identified as greeting. For Participant 5, neither the targeted greetings nor the other social initiations occurred in the postintervention phase.

Maintenance

The performance for the maintenance probes in the classroom for Participant 1 was 0%. From the postintervention phase, there was a descending level change from 100% to 0%. The average performance for the maintenance probes in both the classroom and the intervention room for Participant 2 was 100%. Participant 3’s average performance for the maintenance probes in the classroom and intervention room were 33% and 100%, respectively. The average performance for the maintenance probes in both the classroom and the intervention room for Participants 4 and 5 was 100%.

Social Validity Questionnaire

The parents of the participants identified a greater observed increase in social interaction with others; however, this was not mirrored by teachers and paraprofessionals. The combined averages across all respondents was 4.0 (range = 3–5) for Participant 1, 1.8 (range = 1–3) for Participant 2, 3.3 (range = 2–4) for Participant 3, 3.0 (range = 1–5) for Participant 4, and 3.3 (range = 2–5) for Participant 5.

Discussion

This study expands the literature on using point-of-view modeling to increase the social behavior of students with ASD. With a vertical analysis of the data, a firm relationship between the three target behaviors of the dependent variable and the intervention was established because these behaviors increased only when the point-of-view videos were applied successively across participants.

When examining consistency of data in the intervention phase across participants, there was an overall pattern of consistency, especially with Participants 1, 3, and 4. The data for Participant 5 were also consistent with the aforementioned participants, with slight variability with the initial practice session probes. Following the first two daily probes, the practice session probes for Participant 2 were highly variable. However, all participants were able to meet the established criterion level.

When examining consistency of data in the postintervention phase across participants, there was considerably less uniformity. The trend in the postintervention phase remained high and stable for Participants 2 and 4. However, the data were increasingly variable for Participant 1, Participant 5, and especially for Participant 3, which led to the implementation of the retraining procedures.

Point-of-View VM and Greetings

Participants, 1, 3, and 4 rapidly met the criterion level. Participants 1 and 4 specifically were more likely to engage in verbalizations (primarily with adults), and had a greater preference for visual stimuli, which was demonstrated by their attention to the video model with limited prompting. In addition, Participants 1, 3, and 4 were responsive to the correction procedures, by rapidly engaging in the targeted behaviors following correction, and reinforcement. It is also suspected that the minimized distractions within the intervention room may have been conducive to imitating the targeted behaviors, whereas, the novelty of the environment was more of a distraction for other participants.

It is possible that these participant characteristics (i.e., verbalizations, preference for visual stimuli, responsiveness to correction, and reinforcement) may be prerequisites for point-of-view VM to be effective. This aligns with studies conducted by Sancho et al. (2010) and Tereshko et al. (2010), who suggested that participants’ attention to a video or television may be a prerequisite skill. In addition, Tetreault and Lerman (2010) identified that participants’ receptive and expressive language may facilitate engagement in more expansive communication. A preference for visual stimuli may lead participants to more attentively view the video model, and responsiveness to correction and reinforcement may enable the learning of and engagement in the targeted behaviors.

Both Participant 2 and Participant 5 also reached the established criterion; however, there were some notable variation. The initial inconsistencies may be due to the sudden change in the classroom routine. In addition, the investigator observed Participant 2 frequently engaged in perseverative speech. Although the behavior decreased throughout the study, it may have interfered with his engaging in the target behaviors more consistently. Similarly, Participant 5’s responding may have been hampered due to perseverations and mimicking of the music in the video clip preceding the video model. However, following the correction procedures, the participant accurately performed the targeted behaviors and immediately received reinforcement with the iPad. Because the iPad was such a positive reinforcer for Participant 5, it is believed the frequency of the targeted behaviors increased over time with this reinforcement in place.

Anecdotally, the investigator observed Participants 4 and 5 verbally stating “Hi” or “Hello” immediately before the voice in the video model stated the greeting to Communication Partner 1. This suggests the participants were actively viewing the video, which led them to memorize and anticipate the modeled behaviors. This engagement in the video may have then aided them in performing the targeted behaviors during the practice session probes. During practice session probes, Participants 4 and 5 were also observed looking directly down at the floor, and then immediately looking up once they heard the intervention room door open and engaging in the greeting. These observations suggest Participants 4 and 5 made a direct connection between the first-person perspective of the video model and what they themselves were to see during the practice session probes.

Social validity questionnaire

In explaining their rating, parents of the participants noted an observed increase in social interaction and communication. Most notably, the parents of Participant 4 observed improvements in greetings and interaction with peers in the community, and also observed their child saying goodbye when leaving. Similarly, the classroom teachers and paraprofessionals observed an increase in greetings, eye contact, and waving for Participants 1 and 4. However, for Participant 2, the classroom staff observed a continued need for prompts for the student to interact with peers. The classroom teacher and paraprofessionals also observed minimal change in interactions with peers and a sustained preference for gaining adult attention for Participants 3 and 5.

It is not clear why the observations provided by parents differed with some observations provided by teachers and paraprofessionals. The investigator frequently communicated with the classroom staff about the progress of each participant, which may have influenced how the classroom staff completed the questionnaire. It is also possible that when compared with Participants 1 and 4, who engaged in more verbalizations primarily with adults, the remaining participants still engaged in less communication.

Fading of Prompting

The fading of prompting led to independent performance of the targeted behaviors and all participants were able to reach the criterion level in the intervention phase. When prompting and reinforcement were both removed during the postintervention probes in the classroom, engagement in the targeted behaviors continued for Participants 2 and 4. Participant 2 would rapidly greet and wave to Communication Partner 1 in the classroom. In comparison with other participants, Participant 4 was more engaged with Communication Partner 1 within the intervention phase, and frequently greeted Communication Partner 1 by his name, smiled, and waved. This increased interaction with the communication partner may have then generalized to the subsequent postintervention phase.

However, the data were highly variable for Participants 1, 3, and 5, who all required retraining. The variable performance may be explained by the videos and songs displayed on the interactive whiteboard in the classroom during the postintervention probes. In addition, the classroom environment was oftentimes active, with increased movement and vocalizations by other students, educators, and paraprofessionals. These factors, which starkly contrasted the environment in the intervention room, may have prevented these participants from hearing the classroom door being opened and then shut, or seeing Communication Partner 1 standing in the classroom. These results and the potential impact of the classroom environment coincide with findings from Hine and Wolery (2006), who also indicated that limited performance of the target behaviors within the classroom may be attributable to the activities within the environment competing for participants’ attention.

A single retraining led to immediate performance of the targeted behaviors for Participant 1. However, this was not true for Participants 3 and 5. It is not clear why Participant 3 required numerous retraining and was unable to reach the established criteria. Satiation of baseline probes may have led Participant 3 to become comfortable with frequently seeing and not responding to Communication Partner 1 entering the classroom. Attention to other visual stimuli, including videos and books, may have played some role in the need for multiple retraining for Participant 5. Similar to Participant 3, satiation of baseline probes, may have led to lower performance within the classroom. In addition, on several occasions in the postintervention phase, Participant 5 would see Communication Partner 1 and walk toward the door and ask to work in the intervention room. These observations suggest Participant 5 may have only associated Communication Partner 1 with the intervention sessions conducted in the intervention room and not in the classroom.

Generalization With Communication Partner 2 and Natural Generalization

In the generalization phase, which involved Communication Partner 2 and was conducted in a novel environment, none of the participants engaged in all three targeted behaviors. Similar to the classroom environment, the environments in which the generalization probes were conducted contained a number of distractions.

Similarly, generalization probes conducted during naturally occurring opportunities for social initiations demonstrated limited generalization. The only increase in frequency of social initiations was measured for Participants 1, 2, and 4. These results reflect the findings by Sancho et al. (2010), in which the investigators found limited engagement in unscripted play actions and vocal scripts and limited generalization following the intervention. The results also align with Tetreault and Lerman (2010), who concluded that there was minimal generalization with the social scripts.

Maintenance

Overall, these results demonstrate that all participants were able to maintain the targeted behaviors in the intervention room. However, the results were variable when maintenance was measured within the classroom. Only Participants 2, 4, and 5 maintained the behaviors in both the classroom environments and the intervention room. The performance of Participants 2 and 4 is consistent with their high performance in the intervention and postintervention phases. For Participant 5, the multiple retraining and, thus, multiple viewings of the video model in postintervention may have aided the participant in maintaining the skills.

For Participant 1, receiving the retraining only once in the postintervention phase may have contributed to the lower performance with the maintenance probes in the classroom. The performance for Participant 3 was consistent with his performance in postintervention, which may be attributable to the amount of distractions within the classroom and satiation of baseline probes.

Strengths of the Study

Methodologically, the study aligned with quality indicators suggested by Kratochwill et al. (2010) and Horner et al. (2012). In addition, the intent of the study was to provide socially important findings, which is specifically emphasized by Horner et al. (2012), who discuss the “feasibility of achieving the effect in typical social contexts” (p. 273). In alignment with applied research practices, the study demonstrates that the intervention may be used by educators within the school context.

Limitations of the Study

The scale of baseline probes conducted may have desensitized the participants to Communication Partner 1 repeatedly entering and exiting classroom without having to respond (i.e., satiation of baseline probes). This possible limitation may have affected the data and overall, may have weakened participants’ performance in the postintervention phase and maintenance phase. This is also suggested by Hine and Wolery (2006), who discussed satiation of the intervention materials as a possible reason for limited generalization in their own study.

In addition, data in the postintervention, generalization (with Communication Partner 2 and naturally occurring opportunities), and maintenance phases were minimal. This may, in part, be due to the primary usage of a single student as Communication Partner 1. It is imperative that individuals with ASD not only perform the skills in the environment in which the intervention was implemented (i.e., under controlled conditions) but also proceed to engage in the behaviors in novel environments and with novel peers, thus demonstrating the learning of a general response class and improved social competence. The minimal generalization of the skills may also be attributable to the praise and reinforcers acting as a discriminative stimulus within the intervention phase, which limits the conclusions that may be drawn regarding the effect of the point-of-view video model separate from other factors that might also be influencing the participants’ behavior.

Implications for Practitioners and Future Research

The findings from the current investigation give rise to future research on point-of-view VM as an intervention targeting social communication and interaction, as well as implications for practitioners. Given the possible satiation of baseline probes and its impact on participants’ overall performance, conducting the study in three tiers, replicated twice or a multiple-probe design, may address the issue. Future research should also focus on generalization of greetings gained through this form of VM. Instead of the passive train and hope approach, Stokes and Baer (1977) provide recommendations for proactively fostering generalization. Potential approaches include sequential modification, natural maintaining contingencies, and train sufficient exemplars. The viewing of the point-of-view video model prior to generalization opportunities, and training other peers to extend the interaction through play or conversation may serve as natural reinforcement that will both generalize and maintain the targeted skill. Other promising methods may include the filming of point-of-view video models involving multiple exemplars that include multiple communication partners within novel, natural environments. Based on knowledge of individual students, practitioners may consider whether video priming or prompting is more appropriate, and whether a video clip is needed to gain the attention of the participant and prepare the participant to view the video model. In addition, a child’s voice may be used to provide any vocal models in the video to further provide authenticity. Practitioners and future researchers may also consider filming multiple point-of-videos in other contexts with increasing degrees of distraction, as an approach to scaffolding generalization of skills in environments with increased visual and auditory stimuli. In conjunction with filming multiple point-of-view video models, more opportunities to train, practice social initiations, and interact with typically developing peers in natural contexts may increase both generalization and maintenance for these participants. In addition, periodically returning to the point-of-view video model as a way to remind and prompt the student may be necessary to promote maintenance. In conjunction with the study conducted by Tetreault and Lerman (2010), it would be beneficial to examine the full extent of the intervention in terms of teaching social skills, including greetings followed by maintained conversations. Furthermore, continued research is needed to explore point-of-view VM in isolation and in comparison with other forms of VM.

Footnotes

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent

Informed consent was obtained from all individual participants’ parents or guardians included in the study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jennifer Lee Kouo

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Baron-Cohen

Leslie

A. M.

Frith

(1985). Does the autistic child have a “theory of mind”? Cognition, 21, 37–46.

Barton

Lawrence

Deurloo

(2012). Individualizing interventions for young children with autism in preschool. Journal of Autism and Developmental Disorders, 42, 1205–1217.

Goldstein

Naglieri

J. A.

(2012). The Autism Spectrum Rating Scale (ASRS). Toronto, Ontario: Multi-health Systems.

Hine

Wolery

(2006). Using point-of-view video modeling to teach play to preschoolers with autism. Topics in Early Childhood Special Education, 26, 83–93.

Horner

R. H.

Swaminathan

Sugai

Smolkowski

(2012). Considerations for the systematic analysis and use of single-case research. Education and Treatment of Children, 35, 269–290.

Jung

Sainato

(2013). Teaching play skills to your children with autism. Journal of Intellectual and Developmental Disability, 38(1), 74–90.

Kratochwill

T. R.

Hitchcock

Horner

R. H.

Levin

J. R.

Odom

S. L.

Rindskopf

D. M.

Shadish

W. R.

(2010). Single-case designs technical documentation. Retrieved from https://ies.ed.gov/ncee/wwc/Docs/ReferenceResources/wwc_scd.pdf

10.

Lee

J. N.

(2015). The effectiveness of point-of-view video modeling as a social skills intervention for children with autism spectrum disorders. Review Journal of Autism and Developmental Disorders, 2, 414–428. doi:10.1007/s40489-015-0061-x

11.

Mason

Davis

Boles

Goodwyn

(2013). Efficacy of point-of-view video modeling: A meta-analysis. Remedial and Special Education, 34, 333–345.

12.

McCoy

Hermansen

(2007). Video modeling for individuals with autism: A review of model types and effects. Education and Treatment of Children, 30, 183–213.

13.

Meisels

S. J.

Marsden

D. B.

Jablon

J. R.

Dorfman

A. B.

Dichtelmiller

(2001). Work Sampling System–4th Edition (WSS). Ann Arbor, MI: Rebus.

14.

National Professional Development Center on Autism Spectrum Disorders. (2014). Evidence-Based Practices. Retrieved from http://autismpdc.fpg.unc.edu/evidence-based-practices

15.

Newborg

(2004). Battelle Development Inventory-2nd Edition (BDI–2). Boston, MA: Houghton Mifflin Harcourt.

16.

Odom

Strain

(1986). A comparison of peer-initiation and teacher-antecedent interventions for promoting reciprocal social interaction and autistic preschoolers. Journal of Applied Behavior Analysis, 19, 59–71.

17.

Rayner

Denholm

Sigafoos

(2009). Video-based intervention for individuals with autism: Key questions that remain unanswered. Research on Autism Spectrum Disorders, 3, 291–303.

18.

Roid

G. H.

Miller

Pomplun

Koch

(2013). Leiter International Performance Scale-Third Edition (Leiter–3). Wood Dale, IL: Stoelting.

19.

Sancho

Sidener

Reeve

(2010). Two variations of video modeling interventions for teaching play skills to children with autism. Education and Treatment of Children, 33, 421–442.

20.

Schopler

Van Bourgondien

M. E.

(2010). The Childhood Autism Rating Scale-Second Edition (CARS–2). Los Angeles, CA: Western Psychological Services.

21.

Shane

Laubscher

Schlosser

Flynn

Sorce

Abramson

(2012). Applying technology to visually support language and communication in individuals with autism spectrum disorders. Journal of Autism and Developmental Disorders, 42, 1228–1235.

22.

Shukla-Mehta

Miller

Callahan

(2012). Evaluating the effectiveness of video instruction on social and communication skills training for children with autism spectrum disorders: A review of the literature. Focus on Autism and Other Developmental Disabilities, 25, 23–36.

23.

Stokes

T. F.

Baer

D. M.

(1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10, 349–367.

24.

Tereshko

MacDonald

Ahearn

(2010). Strategies for teaching children with autism to imitate response chains using video modeling. Research in Autism Spectrum Disorders, 4, 479–489.

25.

Tetreault

Lerman

(2010). Teaching social skills to children with autism using point-of-view video modeling. Education and Treatment of Children, 33, 395–419.

26.

White

Keonig

Scahill

(2007). Social skills development in children with autism spectrum disorders: A review of the intervention research. Journal of Autism and Developmental Disorders, 37, 1858–1868.