Abstract
Efficient vocational skills instruction is needed to meet the needs of a growing number of job seekers with intellectual and developmental disabilities (IDD). This study examined the effects of self-directed video prompting used to teach transition-age students with IDD chained vocational tasks. A multiple probe design across behaviors was replicated across two students. Both students mastered iPhone navigation after observing a model and completing two to three sessions in a training phase. After training, students independently navigated the iPhone and played video prompts. Both students made substantial progress on all three tasks, reaching 100% accuracy in two of the tasks.
The U.S. Congress passed the Individuals With Disabilities Education Act (IDEA; 2004), designating education as the means to establish a “national policy of ensuring equality of opportunity, full participation, independent living, and economic self-sufficiency for all individuals with disabilities” (Section 601). Paid employment after secondary school is an essential step toward realizing the intent of the law that governs special education, as it is inextricably linked to all the desired outcomes listed above. Employment leads young adults, with and without disabilities, toward independence and economic freedom (Kiernan, Hoff, Freeze, & Mank, 2011). Employment is also tied to an individual’s identity, social circle, and degree of community integration (Kiernan et al., 2011).
According to the Human Services Research Institute and National Association of State Directors of Developmental Disabilities Services (2015), a data compilation project of public developmental disabilities agencies across 46 states, students with intellectual and developmental disabilities (IDD) who receive at least one service (other than case management) consistently have relatively low rates of competitive, paid employment once they leave secondary school. In 2015, the employment-population rate for people without disabilities in the United States was 65%, but only 20% of individuals with developmental disabilities between the ages of 18 and 34 indicated they currently had community employment. However, 59% of individuals in this age range without a job in the community reported they would like one (Human Services Research Institute & National Association of State Directors of Developmental Disabilities Services, 2015; United States Department of Labor, 2016). There are several potential reasons for this disparity, including employment discrimination in certain fields (Houtenville & Kalargyrou, 2015), lack of a marketable skill set, social skill deficits, insufficient available supports (Test, Smith, & Carter, 2014), low expectations from family and educators (Doren, Gau, & Lindstrom, 2012), and lack of work and internship experiences (Carter et al., 2016).
Currently, vocational rehabilitation (VR) agencies and transition coordinators in school districts are tasked with preparing every person with IDD to obtain the skills and strategies that will lead to paid employment (IDEA, 2004; Workforce Innovation Opportunity Act [WIOA], 2014). The U.S. Department of Labor and 46 states have taken a philosophical or legislative approach that prioritizes employment in the community as the first option for individuals with disabilities; this approach is called “Employment First” (Association of People Supporting Employment First, 2013). Despite more than 25 years of culture shifts, advocacy, and policy changes, this Employment First movement, led by self-advocates, activists, and supportive organizations, has not yet reached the goal of total systems change, as is evident from the labor force participation rate flat-lining for individuals with IDD (Association of People Supporting Employment First, 2013; Kiernan et al., 2011). Adjustments are still needed in job development, employee training and support, and effective resource allocation to reach the goals of the Employment First movement (Kiernan et al., 2011). To counter the climate of limited resources needed for 1:1 job coaching support in schools and the community, school transition professionals and VR services need tools to teach a diverse set of skills efficiently, with as little staff support as possible. It is possible that many job seekers could self-direct their instruction and long-term supports if provided with a portable model for their vocational tasks. Self-directed instruction would reduce reliance on staff and could be a more efficient approach in some cases.
Portable, individualized instruction and support through technology could be one strategy to supplement the work of job coaches, teachers, and educational aides, as they are often supporting multiple students. Video technology-based instruction has demonstrated growing support in the special education literature over the last 30 years (Banda, Dogoe, & Matuszny, 2011). Video technology has been used to teach a wide range of skills including academic skills such as fraction problem solving (Yakubova, Hughes, & Hornberger, 2015), some vocational and prevocational skills (Bereznak, Ayres, Mechling, & Alexander, 2012; Cullen, Alber-Morgan, Simmons-Reed, & Izzo, 2017), daily living skills such as cooking and cleaning (Shipley-Benamou, Lutzker, & Taubman, 2002; Sigafoos et al., 2007), and fire safety (Tiong, Blampied, & Le Grice, 1992).
Video prompting is a technology-assisted form of simultaneous prompting, in which the model is broken into video clips of each step of the task analysis, so a learner can complete steps of the task as they watch the video clips (Le Grice & Blampied, 1994). Video-recorded models have several advantages over in vivo prompting, because videos provide a consistent model that contribute to consistent results and videos are transportable across settings and staff (Charlop-Christy, Le, & Freeman, 2000; Mechling, 2005). Video clips also have the potential to gain the attention of students who are reinforced by interaction with the video (Charlop-Christy et al., 2000). The use of video prompting is cost-efficient because the video clips can be recorded and displayed on smart phones that the majority of American adults already own (Anderson, 2015). Test et al.’s (2009) review of evidence-based practices in secondary transition cited job training using computer-based instruction, including video and audio-visual self-prompting strategies, as being a strategy for skill-building with transition students development supported with moderate evidence.
Similar to video prompting, self-directed video prompting is a form of simultaneous prompting in which a model is broken down into a series of video clips. Self-directed video prompting is distinct from video prompting, because learners independently navigate through short video clips of individual steps of a task analysis, as they are completing the task (Cannella-Malone, Brooks, & Tullis, 2013). Self-directed video prompting is an instructional strategy that has potential to lend itself to efficient skill building for transitioning students and young adults in vocational settings. Video prompts operated by the student add an additional advantage of cost-efficiency because they decrease the need for one-on-one staff support for prompting tasks not yet been mastered, allowing instructors to work with multiple students at the same time (Mechling, 2005). Self-directed video prompting has the potential to be discrete in some work settings where an individual would stand out more with a job coach than they would occasionally referencing a mobile device, potentially decreasing stigma while still providing lasting support. However, this procedure still involves a small degree of job coach intervention during acquisition, as some students may need occasional error correction to prevent them from practicing errors.
Recent studies investigating error correction procedures have demonstrated promising effects for use with video prompting. Video-based error correction procedures, as Cannella-Malone, Wheaton, Wu, Tullis, and Park (2012) used in their video prompting study, typically involve informing students that their response was incorrect and playing video clips a second time following an error. Video prompting studies have varied the timing of implementation of error correction procedures, with some adding the procedure after the individual’s acquisition trend stabilizes (Goodson, Sigafoos, O’Reilly, Cannella, & Lancioni, 2007), and others using error correction from the outset with self-directed video prompting (Van Laarhoven, Johnson, Van Laarhoven-Myers, Grider, & Grider, 2009). Cannella-Malone et al. (2012) used an alternating treatments design to compare the efficiency of acquisition between conditions of video prompting with error correction from the beginning of intervention and error correction added after plateau. Results indicated using video-based error correction procedures from the outset led to more efficient acquisition with video prompting suggesting the need for additional research on error correction procedures.
A variety of applications and software have been used to shoot, edit, and display videos in video prompting interventions. Cullen et al. (2017) used the MyPicsTalk app, Cannella-Malone et al. (2013) used an app called inPromptu, and Cannella-Malone et al. (2012) used iMovie to develop their videos. There is a learning curve for teachers for keeping up with each new software, along with idiosyncratic errors in these apps. These complications can be avoided by using widely used commercial technology (Van Laarhoven et al., 2009) that is more likely to be familiar to teachers and students and it is updated frequently to avoid bugs.
Research with self-directed video prompting in vocational settings has grown along with the demand for vocational instruction in recent years (Bereznak et al., 2012; Cullen et al., 2017). These studies have shown relatively rapid acquisition rates, as well as unique challenges to be considered in future replications. Cullen et al. (2017) used a multiple probe across behaviors design to demonstrate the use of self-directed video prompting for the rapid acquisition of vocational skills. Students were trained to access short video clips on an iPad as they completed tasks in their internships. However, this study ran into some logistical concerns inherent in attempting to implement in an integrated employment setting, such as having to use in vivo prompts to complete some of the tasks in the baseline phase to meet the demands of the integrated environment, despite the possible learning effect on the student (Cullen et al., 2017). Few self-directed video prompting studies have gathered and reported data on student’s accuracy in navigating the technology (Cannella-Malone et al., 2013; Cullen et al., 2017). Navigation data are an important measure in studies claiming to be student-directed, so the data can demonstrate whether the student’s technology use was sufficiently independent to be considered self-directed.
Given the continued need for increasing opportunities in vocational settings for individuals with IDD, this study sought to answer the following research question:
This study will expand on the previous work of Cullen et al. (2017) and Bereznak et al. (2012) by increasing the efficiency of the preparation procedures by eliminating extraneous software or video editing processes to avoid teachers having to either buy and learn a new app, such as the MyPicsTalk app in the Cullen et al. (2017) study, or deal with infrequent updates and intrusive advertisements on free apps. This study will also increase the amount of experimental control over tasks by opting for a school-based training setting and novel tasks for the students.
Method
Students
Two students were included in this study. The school district’s transition program teacher was asked to recommend students between 18 and 22 years of age with moderate to intensive IDD who had the dexterity to operate an iPhone and needed job training in discrete vocational skills. Student assent and parental permission were obtained before the study began.
Hannah
Hannah was a 21-year-old female with diagnoses of moderate intellectual disability, Down syndrome, selective mutism, and moderate postlingual hearing loss. She was in her third year in the vocational skills transition program. Hannah used an iPad regularly as a communication device. Hannah scored proficient in all academic subjects on her alternate assessments before entering the vocational program. She displayed functional reading and comprehension skills in the classroom by reading and responding to short emails, writing reflections in paragraphs, and answering questions about what she read. She earned a 426 (proficient) reading score on her most recent standardized tests. She attended internship sites at retail shops, restaurants, offices, and animal shelters through the school program, and worked a paid housekeeping internship through her church. Her teacher reported Hannah was working on self-monitoring during chained tasks, completing tasks efficiently, having appropriate social interactions, and asking for help when needed.
Will
Will was a 21-year-old male with diagnoses of moderate intellectual disability and Down syndrome. He was in his third year in the vocational skills transition program. Will scored proficient in all academic subjects on his alternate assessment before entering the transition program. He displayed functional reading and comprehension skills in the classroom by completing vocational skills based on a written schedule, and reading and responding to emails. He earned a 426 (proficient) in reading on his most recent standardized test. Will was interning at a mail room, as well as attending internship sites at retail shops, restaurants, offices, and animal shelters through the school program. Will’s teacher reported he was working on building skills that would allow him to gain responsibility and become more independent in vocational settings, and ultimately find a job that he enjoyed. Will had no communication device and used an android smartphone, so the interface was slightly less familiar to him.
Settings
All iPhone training, baseline, and intervention sessions took place in the students’ transition program homeroom or in the union common space. Both rooms are located within a small, alternative high school in a suburban public school district. The school offered specified programs for students with and without disabilities, such as a specialized science, technology, engineering, and mathematics (STEM) program. The transition program was for students with intellectual disabilities, ages 18 to 22, who defer graduation to obtain additional vocational experience and independent living skills. The homeroom had six large tables, each with four chairs. Student daily and weekly schedules were posted on the wall, and various areas around the room were designated for in-school businesses (e.g., coffee delivery service, snack cart). The 10 students in the transition program moved in and out of this space, as they get ready to go to internship placements and do vocational tasks throughout the school, so there were typically only three to four students in the space at a time. The union common space had tables and chairs, couches, and standing tables. It is a multiuse space where the students ate lunch, socialized, and waited for their next class. This space typically had zero to five students when used for experimental sessions. Experimental sessions were held in the classroom on days when there was sufficient table space and in the union on days when most of the students in the classroom were working on laptops and using the table space.
Materials
iPhone
An iPhone SE running iOS 10.3.3 was used to display video prompts. The iPhone was set up without a password so it could be opened easily.
Videos
Video clips were shot in sequential order using the default Apple Camera app, and video clips were placed into folders by task in the default Apple Photos app. Each video clip began with a verbal instruction followed by a visual demonstration of the step. All video clips were shot from a first-person perspective, over the shoulder of an adult male actor, who was a contact of the first author, unaffiliated with the high school or university. Verbal instructions in the video clips were given by the adult female researcher.
Task analyses for video clips were developed through observation of several special education master’s students completing the three selected tasks. Tasks were selected in collaboration with the classroom teacher based on their appropriateness for student skill level, suitability to portray in video clips, and novelty to students. The tasks ranged between 17 and 40 steps, but were comparable in difficulty because the longer tasks had multiple repeated steps. Full task analyses for all tasks are included in Table 1. The coffee task had 17 video clips that averaged 9 s per clip (e.g., Pull two coffee cups out of supply bin). The name tags task had 28 video clips that were averaged 11 s per clip (e.g., Gently pull up on the middle of the string). The wrapping task had 40 video clips that averaged 10 s per clip (e.g., fold the end of the paper up until it can fold over the top of the box). When one of the video clips was found to be unclear, and a student was making a repeated error as a result, it was replaced with a new video clip that showed the same step from a different angle.
Task Analyses.
Task materials
During iPhone training, both students were provided a three-hole punch machine and a stack of 25 pieces of printer paper. For the name tags task, students were provided with an Avery Top-Loading Hanging Name Badge kit and a small clear container of finished name tags. Materials provided for the coffee task included a tray with three sections and a supply bin containing disposable coffee, coffee cups, lids, stirring straws, sugar packets, creamer, coffee filters, and welcome cards. For the wrapping task, students were provided with an empty box, a roll of wrapping paper, invisible tape, a pencil, and scissors.
Variables and Data Collection
The primary dependent variable was accuracy of task completion defined as completing the step as described in the task analysis and measured as percentage of steps completed correctly. The primary dependent variable was used to make decisions about when to move students into the next phase of intervention.
Data on training phase were collected on a task analysis for the training task (i.e., hole-punching papers). Next to each step, data collectors indicated the student’s level of performance. Two types of data on correct responding were collected. A correct response was also coded if the student completed the step correctly after watching the video clip. Steps completed out of order were counted as correct if the outcome was functionally equivalent. For example, some steps in wrapping can be interchanged without altering the finished product, such as starting to fold on the right or left corner. In this case, observers were instructed to score for the step that the student completed. However, some steps cannot be interchanged without consequence, such as taping before folding. Incorrect responses were coded if the student responded correctly after watching the video clip a second time, or if the student made two errors on the same step and the researcher completed it as a model. Responses that took longer than 4 s to initiate, missing steps, and incorrect responses were also counted as errors. Percentage of steps completed correctly for both the vocational skill were calculated as accurate step completion divided by the total number of steps multiplied by 100.
A secondary measure, accuracy of iPhone navigation, was collected throughout the study to show the degree to which the intervention was “self-directed” in practice. For use of the iPhone, data collectors scored whether the student completed the steps correctly with no prompts, correctly with a verbal prompt, correctly with a gestural prompt, or correctly with a physical prompt. Percentage of steps completed correctly for iPhone use were calculated as accurate step completion divided by the total number of steps multiplied by 100.
Interobserver Agreement (IOA)
IOA and procedural integrity data were collected live with a second in-person observer. IOA data were collected for Will and Hannah, respectively, in 29.4% and 26.3% of baseline sessions, 25% and 29.4% of intervention sessions, and 50% of maintenance sessions for both. One of the two second observers were present for these sessions. IOA was calculated as agreements divided by agreements plus disagreements multiplied by 100. Two doctoral students in special education served as secondary data collectors. The first author trained them by demonstrating how to record on the data sheet, by giving examples of what would be counted as correct and nonexamples that showed what would be counted as incorrect, and being available to answer questions. Overall point-by-point IOA for sessions with Will and Hannah was 100% for all conditions.
Experimental Design
This study used a multiple probe design across three behaviors per student. Baseline data were collected for all behaviors and the first behavior was moved into intervention following at least five sessions showing a stable flat trend. After meeting the mastery criterion of at least five sessions demonstrating a clear effect and a stable or increasing trend for all three intervention tiers, reaching a minimum performance criteria of at least 95% accurate task completion, students were moved into the maintenance phase.
Procedures
The authors of this study were a master’s student, a doctoral candidate, and a professor in a special education program.
iPhone training phase
Experimenters selected hole punching as a task in the iPhone training to assess the student’s competence with technology use in isolation. Students were individually presented with an in vivo model of how to use the iPhone in which the experimenter demonstrated (a) waking up the device, (b) opening the pictures app, (c) selecting the folder from within the app that matched the task (e.g., “hole punching”), (d) playing the first video clip for that task, (e) doing the step shown in the first video clip, and (f) navigating between video clips. Each student was asked to “Hole punch these papers by following the instructions on the iPhone.” If a student did not independently initiate waking the device within 4 s, or if the student made an error in either iPhone navigation or the task, the researcher stopped the student and used a least-to-most prompting hierarchy until the student demonstrated the correct action. At the end of each iPhone training session, students received nonspecific praise (e.g., “thanks for working with us”). Students were required to reach 90% accuracy for both iPhone use and completion of the task before moving into baseline.
Baseline
Students were provided with a verbal direction to complete their novel vocational task and a paper copy of the task analysis so the steps were available to them, but the iPhone was not available to them. Experimenters used a multiple opportunity method during baseline sessions. When a student made an error, the researcher interrupted the student, stood in front of the student to block their view of the task, and completed the step. Then, the experimenter gave the student an opportunity to complete the next step. No verbal or gestural prompts were given to the student at this time. Students were thanked for their participation after each baseline session.
Self-directed video prompting
Students were given the iPhone and instructed to use the video clips to complete the task. If the student did not initiate the iPhone within 4 s, or if after the video clip the student made an error, the error correction process began. The first time there was an error, the researcher interrupted the student saying, “that’s not quite right,” reset the scene, gave the student specific verbal feedback about the error, and verbally prompted the student to rewatch the video clip and pay close attention to the part they missed (e.g., “You took too many sugar packets. Play the video clip one more time and look closely at how many sugar packets are needed” or “you didn’t start folding the corner of the paper, play the video clip again and give it a try”). If the student made an error or did not initiate within 4 s during the second attempt, the experimenter completed the step for the student as a model. After each step the student completed correctly, they would move on to the next video clip and the process was repeated. Prior to moving on to the next skill, each student’s data had to show a clear and increasing trend across a minimum of three trials.
Maintenance
Maintenance probes were taken every 2 weeks for three sessions, which began immediately after the student mastered all three tasks. During maintenance probes, the iPhone was placed in a location visible to the students, but they were not directly instructed to use it, and no error correction procedures were implemented. Following maintenance, a probe was conducted that mimicked baseline procedures, removing access to the video clips.
Procedural Integrity
Procedural integrity data were collected for Will and Hannah, respectively, in 29.4% and 26.3% of baseline sessions, 25% and 29.4% of intervention sessions, and 50% of maintenance sessions for both. Procedural integrity data were recorded by a second data collector and calculated as the number of applicable steps completed correctly divided by the total number of applicable steps multiplied by 100. Data were collected on a fidelity checklist to indicate whether the experimenter (a) provided the student with needed materials, (b) verbally directed the student to use the iPhone to complete the task, and (c) completed error correction procedure using some or all three steps in correct sequence (e.g., specific feedback, replaying the video clip, modeling the step). Overall point-by-point procedural integrity for sessions with Will was 100% accurate completion for all conditions. Procedural integrity for Hannah was 100% for baseline conditions, 96.4% (range = 85.7%–100%) for treatment conditions, and 100% for maintenance probes.
Social Validity
Social validity data were collected on procedures and outcomes. After maintenance, students were given the opportunity to express their opinions about the intervention by answering a few short questions with three options, “yes,” “maybe,” and “no.” The researcher also asked the teacher five open-ended questions regarding her perception of the effectiveness of the intervention and its viability in a transition program setting.
Results
Results indicated an increase in accuracy of task completion coincided with the use of self-directed video prompting for both students across all three tasks. Acquisition data showed a similar pattern across tasks and was replicated across students. Both students demonstrated accurate use of the iPhone throughout all tasks in the study.
iPhone Training
Hannah
After viewing an in vivo model once, Hannah completed 87.5% of the steps for iPhone navigation accurately, and 100% of the hole-punching steps accurately. In the first training phase session, Hannah required a verbal prompt to unlock the iPhone and to select the correct task. She completed two sessions of training before reaching 100%. Once self-directed video prompting began, Hannah used the iPhone correctly in 100% of the intervention sessions.
Will
After viewing an in vivo model once, Will completed 81% of the steps for iPhone navigation accurately and 100% of the hole-punching steps accurately. In the first training phase session, Will required a verbal prompt to select the correct app, select the task within the app, and to play the first video clip. He completed iPhone training in three sessions before reaching 100%. Once self-directed video prompting began, Will used the iPhone correctly in 100% of the intervention sessions.
Accuracy of Task Completion
Hannah
Results for Hannah’s task accuracy are shown in Figure 1. Hannah did not receive any “independent plus” scores because she did not complete any steps prior to playing the video clips. During baseline, Hannah did not complete any steps of the three tasks. In her first session of self-directed video prompting, Hannah accurately completed 89% of name tag assembly steps. Hannah’s name tag assembly accuracy increased to 100% by the second session and remained at 100% for the remaining intervention sessions. Hannah’s average accuracy during name tag assembly intervention was 97.8% (range = 89%–100%). Hannah made one error in threading the string backward in the first session of intervention for this task, requiring a Level 1 error correction (i.e., provide specific feedback, reset the scene, direct to replay video clip). In maintenance sessions, Hannah’s performance remained at 100% accuracy. Seven weeks after the last intervention session, a probe was conducted without the video prompts, and Hannah’s performance dropped to 28.4%.
Hannah.
In the first session with self-directed video prompting, Hannah accurately completed 88% of the steps for the coffee trays task, which increased to 100% by the second session and remained at 100% for all remaining intervention sessions. Hannah’s average accuracy for the coffee tray task was 98% for all intervention sessions (range = 88%–100%). Both errors Hannah made with this task required Level 1 error correction. In maintenance sessions, Hannah’s performance remained at 100% accuracy. Seven weeks after the last intervention session, a probe was conducted without the video prompts, and her performance dropped to 64% accuracy.
In the first session of self-directed video prompting intervention, accuracy for the third task (wrapping) increased to 92%. Hannah demonstrated improvement from baseline sessions, but did not reach 100% accuracy with this task because she struggled with centering the box on the paper, often centering the box on the width or the length of the paper, but not both. Average accuracy was 96% for all intervention sessions (range = 92%–97%). Both errors Hannah made with this task required Level 1 error correction. In maintenance sessions, Hannah’s performance remained at 96% accuracy. Seven weeks after the last intervention session, a probe was conducted without the video prompts and her performance dropped to 17.5% accuracy.
Will
Results for Will’s task accuracy are shown in Figure 2. Will did not receive any “independent plus” scores because he did not complete any steps prior to playing the video clips. During Baseline, Will did not complete any of the steps of the three tasks independently. In the first session of self-directed video prompting, accuracy for name tag assembly increased to 89.3%. Accuracy reached 100% in the following session and remained at 100% for all remaining intervention sessions. Will’s average accuracy during intervention was 98.2% (range = 89.3%–100%). All three errors Will made with this task required Level 1 error correction. In maintenance sessions, Will’s performance remained at 100% accuracy. Eight weeks after the last intervention session, a probe was conducted without the video prompts, and his performance dropped to 28.4%.
Will.
In the first session of intervention, accuracy for the coffee trays task increased to 94.1%. Accuracy remained at 94.1% for three more sessions until one of the videos for this task was reshot at a different angle. After the new video was in place, performance reached 100% for all remaining sessions. Will’s average performance on the coffee trays task was 96% (range = 94.1%–100%). All four errors Will made in this task required Level 1 error correction. Will’s performance remained 100% accuracy in maintenance sessions. Eight weeks after the last intervention session, a probe was conducted without the video prompts, and his performance dropped to 52.9%.
In the first session of intervention, accuracy for the wrapping task increased to 90%. Will made significant gains from baseline, but did not reach 100% accuracy for this task because he had ongoing struggles with some of the fine motor skills required to stick the tape in the correct location and accurately fold on the first try. Average accuracy in intervention sessions was 91.5% (range = 90%–92.5%). Out of the 17 errors for this task, 94% required Level 1 error correction, and 6% required Level 2 error correction (i.e., reset the scene, provide specific feedback, and complete the step as a model). In maintenance sessions, Will’s performance remained at 92.5% accuracy. Eight weeks after the last intervention session, a probe was conducted without the video prompts, and his performance dropped to 25% accuracy.
Social Validity
Both students responded that they enjoyed using video clips to learn the three tasks for this study and that they would like to use the iPhone video clips to learn something else. Will indicated “yes,” he would feel comfortable using the video prompts in an integrated work setting, but Hannah indicated that “maybe” she would be comfortable.
After reviewing the data and observing in-class sessions, the teacher indicated that she thought students satisfactorily learned the target skills. When asked what she liked about the procedure, she responded she liked how it held the attention of students who struggled to focus on longer tasks, saying that it was a “night and day” difference in focus for one of her students. When asked if this procedure would be socially acceptable in the internship settings her students work in, she described how it would work better in some settings than others, but that she saw potential for students learning office and retail skills. Anecdotally, the teacher expressed enthusiasm for the intervention during and after the study, and shared self-directed video prompting lessons she created for other students in the classroom.
Discussion
This study investigated the effects of self-directed video prompting on the accurate completion of chained vocational tasks by young adults with moderate intellectual disabilities. This study used a brief training in iPhone use with least-to-most prompting, self-directed video prompts viewed on an iPhone, and a video-based error correction procedure combined with specific feedback. Both students’ results indicated a functional relation between self-directed video prompting and accurate task completion.
Self-directed video prompting was not only effective in teaching student’s novel skills but it also had strong social validity in this transition program setting for continued, and expanded, use. After observing the first few sessions of the study, the teacher expressed interest in learning to make video clips to teach some of the other students in her class skills, such as shoe tying. The teacher and educational aides expressed enthusiasm about the students’ progress and implemented self-directed video prompting with other students throughout the course of this study. By the end of the study, the teacher was designing a system of Quick Response (QR) codes printed and attached to task materials that would link to video prompts, to simplify navigation through large numbers of video clips. This continued and expanded use of the intervention lends social validity to the intervention’s usefulness and feasibility in a transition program setting with a range of students with moderate IDD.
This study extends the literature on self-directed video prompting in several ways. First, the present findings contribute to the literature by replicating previous studies suggesting video prompting is a viable instructional strategy for acquiring chained tasks (e.g., Cannella-Malone et al., 2006; Sigafoos et al., 2007). Results of this study showed an immediate and consistent effect on the target behavior, similar to other studies that have targeted vocational and daily living skills using self-directed video prompts (e.g., Bereznak et al., 2012; Cannella-Malone et al., 2013). However, it should be noted that similar self-directed video prompting studies by Payne, Cannella-Malone, Tullis, and Sabielny (2012) and Cullen et al. (2017) found significantly more variability in their results between sessions and across students, yet our findings suggest a high degree of consistency across tasks and students. Potential explanations for this are difficult to extrapolate from the small number of relatively homogeneous students included in this study. However, given these students’ extensive experience with technology through their phones, iPads, and laptops at school, their reinforcement history with technology may have positively biased their willingness to engage with the technology.
Second, this study showed a dramatic drop off in performance when the video was removed during maintenance. A handful of video prompting studies have probed performance without videos with mixed results. Some studies, such as Graves, Collins, Schuster, and Kleinert (2005), showed high levels of maintenance without the videos, although other studies, such as Sigafoos et al. (2007), showed results similar to ours, showing a drop in performance without video access. However, this study implemented a “chunking” procedure in which video clips were combined to systematically reduce the number of prompts needed. Third, relatively few studies have collected data on the student’s navigation of the technology (Cannella-Malone et al., 2013; Cullen et al., 2017; Payne et al., 2012). Results of this study support the findings of Cullen et al. (2017) that brief training in technology use can be sufficient to learn to access technology-based instruction. Cullen et al. (2017) included students with similar traits (e.g., age, disability, socioeconomic status) to students included in this study, which may account for the similar results. Contrary to these findings, other studies have had less promising outcomes for their student navigation data. One of the students in Payne et al. (2012) did not learn to navigate the iPod Touch independently during the course of the study. This inconsistency suggests variables such as the student’s previous reinforcement history with the technology and the severity of disability, may affect navigation outcomes.
Implications for Practice
Findings from this study have implications for practice in special education transition programs and VR. The results of this study showed students can become dependent on the videos, unable to complete the task without them. Mechling, Gast, and Seid (2009) used Personal Digital Assistant technology to provide prompt options for students, allowing them to choose between still images, still images with voice over, and video clip with voice over, so that each student receives the amount of prompting they need, without adding unnecessary stimuli in their prompts. This strategy could serve as a model for practitioners because it contributed to an accessible learning structure for a variety of students and facilitated fading.
Literature on video prompting has reported success across many tasks, but the procedure is most often applied to chained tasks that are consistent observable skills (Banda et al., 2011). The findings of this study suggest other considerations for practitioners when selecting tasks to teach using self-directed video prompting. The students in this study struggled with some tasks more than others due to their initial levels of fine motor skills. Therefore, it is important to consider prerequisite skill levels in areas such as fine motor skills to maximize student success. If possible, practitioners should consider consulting with an occupational therapist to select appropriate tasks or modify existing tasks to suit the student’s skill level.
The video prompts used in this study had visual and audio components. After a few sessions of intervention, Hannah stopped looking at the video clips, and would press play and look away from the video clip at the floor or the wall. Hannah continued to accurately complete the task without looking at the visual prompt, presumably relying on her memory and auditory cues. Unlike Hannah, Will seemed to rely almost exclusively on visual cues in the video clips. Teachers should consider using video prompts with narration to provide instruction through multiple sensory pathways. This might create increased opportunity for success because students can use whatever format they prefer and it creates a built-in accommodation for students with mild to moderate hearing or vision impairment.
Both students in this study used touch screen technology in their daily lives. Some learners may require significantly more intensive instruction to learn to use the technology. In cases in which the job seeker lacks the basic navigation prerequisite skills, trainers should evaluate their instructional priorities, as self-directed video prompting may not be the most efficient tool for all students.
Hannah completed her tasks much more slowly than would be acceptable to meet the standards of most competitive workplaces. In such cases, to complete the task in a more competitive amount of time, teachers may consider pairing self-directed video prompting with differential reinforcement of higher rate procedures.
Limitations and Suggestions for Future Research
This study had several limitations that should be considered when interpreting the results. First, this study was conducted with students with a relatively high level of experience with touch screen technology and navigation through tablets and phones. If navigation of similar technology was a novel skill, students may need more intensive instruction to proficiently navigate the device independently prior to intervention. Second, given that students were still receiving services through IDEA, in a school-based program, this study was conducted in a school rather than a competitive work setting. Future studies should consider extending this study by testing generalization to an integrated work setting. However, it is important to consider the challenges of maintaining experimental control in a study in an integrated work setting. To complete a video prompting study in a competitive work environment, future researchers would need to have high levels of employer buy-in, some scheduling flexibility so that targeted tasks are only trained during experimental sessions, and a work environment that can tolerate baseline measures. Third, this study included two students, both of whom had a moderate intellectual disability and Down syndrome. To validate these results for broader populations, future researchers should attempt to include more diverse groups of students and students with more intensive support needs.
Fourth, this study did not complete a pre-study fine motor assessment, and as a result, one student was unable to achieve 100% independence for one of the vocational tasks. Future research should assess motor skills and suggest appropriate accommodations for the target skills. Fifth, this study did not maintain without the videos after intervention. Future research could also promote decreased staff reliance by pairing self-directed video prompting with training in how to recognize errors and recruit trainer attention to eliminate the need for someone to be constantly watching for them to do error correction, increasing efficiency and independence. Sixth, this study did limited training with the second observers, and did not assess for accuracy before the study. Accuracy of the second observers was not tested throughout to prevent observer drift.
Footnotes
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.