Empowering Young Adults With Significant Disabilities to Learn New Employment Skills Through Self-Directed Video-Based Instruction

Abstract

Video-based instruction (VBI) is an effective means to teach employment skills to young adults with intellectual and developmental disabilities (IDD). Within the category of VBI, video prompting (VP) is well-suited to teach multistep routines, and video modeling (VM) is better suited for social interactions. In this multiple probe design, three young adults with IDD in the U.S. Midwest were taught to make badges with self-directed VP with error correction, and to request help for missing materials with VM. Functional relations were demonstrated between VP and progress on the badge-making skill, and VM and progress on requesting help. These findings demonstrate the utility of combining VP and VM to address both multistep routines and social skills in the workplace. We offer practical implications for how teachers and job coaches can leverage VBI to teach different kinds of employment skills to people with IDD.

Keywords

employment skills self-directed video prompting video modeling self-fading behavior significant disabilities

Employment is integrally related to one’s financial independence and subjective well-being (Blustein, 2008; Buttler, 2022). However, employment is often an elusive goal for adults with significant disabilities. The employment rate for adults with significant disabilities—most recently estimated to be 15%—has persistently lagged behind the employment rate for adults without disabilities (National Core Indicators, 2022). Barriers to employment for this group of individuals include their lack of employment-related skills and external realities (e.g., race and gender prejudice, organizational inflexibility, employer expectations, and lack of employment support and services; Anderson et al., 2021; Shepard et al., 2020).

Policymakers, researchers, and community support personnel have been advocating for system-level employment support (e.g., integrated employment environment, work benefits, employment service delivery) for individuals with significant disabilities (Nord et al., 2013). While these system-level changes are critically important, it is equally important that individuals with significant disabilities are well prepared for employment in terms of technical skills and soft skills. Technical skills involve the knowledge and ability to complete work tasks. In contrast, soft skills refer to “skills, abilities, and traits that pertain to personality, attitude, and behavior rather than to formal or technical knowledge” (Moss & Tilly, 1996, p. 327). These include skills such as communication, courtesy, flexibility, interpersonal skills, and professionalism. Both technical skills and soft skills are critically important to success in the workplace (Robles, 2012).

There is substantial research literature focused on developing and testing interventions to improve these employment skills. Early work in 1970s and 1980s started with instructional approaches based on applied behavior analysis and direct instruction methods. These methods involved breaking down a work routine into its component parts, providing visual, verbal, or physical prompts to help people with significant disabilities to learn the steps, and then systematically fading prompts over time (Nord et al., 2013). Later on, self-management strategies (e.g., self-delivered prompts, self-monitoring, self-recording, self-delivered consequences; Browder & Shapiro, 1985) were added to promote working independence with their unique features on improving individuals’ self-awareness and behavior management, fading unnecessary external prompt, and promoting skill acquisition and work efficiency (Barczak & Cannella-Malone, 2022; Cihak et al., 2007; Hughes et al., 2006; Storey, 2007). More recently, technology-based interventions (e.g., video-based instruction [VBI]) have emerged with advantages such as flexible fading procedures, less reliance on external prompts, and reduction of physical demands in some jobs (Nord et al., 2013; Sun & Brock, 2023b).

Video Modeling and Video Prompting

Using video footage to demonstrate a target task, VBI is one of the most frequently used technology-based interventions in research studies (Gilson et al., 2017). Research has extensively examined two major formats of VBI: video modeling (VM) and video prompting (VP; Park et al., 2020; Sun & Brock, 2023b). The difference between the two approaches is that VM involves the learner watching a video of the entire skill and then completing the entire skill, while VP involves watching a short video clip of the first discrete step, completing that step, and then watching a video clip of the subsequent step (Cannella-Malone et al., 2006).

Researchers have suggested that segmented videos could benefit procedural learning (i.e., learning to perform a series of steps to achieve a particular goal; Card et al., 1983) and reduce cognitive load (Biard et al., 2018). This is consistent with findings that VP may be a more effective approach than VM, especially when teaching students with moderate and severe intellectual disabilities who may struggle with working memory (Cannella-Malone et al., 2011; Mechling et al., 2014). Furthermore, VP has been shown to be a versatile approach, with researchers reporting effects across academic, daily living, and employment skills (Cannella-Malone et al., 2011; Kellems et al., 2021; Sun & Brock, 2023b). In addition, there is evidence that individuals with significant disabilities can be taught to manipulate technology to access video prompts independently. This approach, called self-directed VP, holds tremendous promise in empowering people with significant disabilities to independently access the support they need to learn new vocational and daily living skills (Bereznak et al., 2012; Heider et al., 2019; Sun & Brock, 2024).

While VP is a very effective approach for many multistep technical skills, it may be less feasible for some soft skills. For example, if teaching social interactions with co-workers, it would be awkward and cumbersome to alternate between social interaction and watching video clips. Video prompting also may not be a good match for teaching learners to recognize when a skill is needed, such as when to ask for help. In these situations, VM may offer a superior approach because it does not interrupt a social interaction and can be used to teach the learner to recognize the situation in which a skill is needed. Indeed, in a recent review of research involving VBI to teach employment skills, the authors found that VM was the most common and effective tool for teaching “soft” skills such as problem-solving skills, job interview skills, and social skills (Sun & Brock, 2023b).

Although research has extensively examined the effectiveness of VM and VP separately, few studies have used these approaches in combination to simultaneously teach technical and soft skills. The features of VP and VM enable VBI to be one of the promising interventions that can simultaneously address comprehensive needs to retain a job, increasing individuals’ competence toward competitive integrated employment (CIE). In addition, more evidence is still needed regarding procedure variations (e.g., error correction or skill generalization) of VBI and individuals’ learning patterns (e.g., how they initiate learning or fade the prompt; how they operate the device). For example, some studies have found self-directed VP with error correction procedures contributed to skill acquisition, and video prompts were important for skill maintenance (Heider et al., 2019; Van Laarhoven et al., 2009; Yakubova et al., 2019). A few studies have found mixed results on device operation performance during or after device training (Cannella-Malone et al., 2013; Cullen et al., 2017; Payne et al., 2012). Furthermore, self-initiated fading behavior has been reported yet to be systematically tracked (Cannella-Malone et al., 2013).

To address these specific gaps in the literature, the current study further examined the following research questions:

Research Question 1: Is there a functional relation between self-directed VP and vocational skill acquisition for secondary students with significant disabilities?

Research Question 2: Is there a functional relation between VM and social skill acquisition for secondary students with significant disabilities?

Research Question 3: To what extent do students with significant disabilities naturally fade their video prompts?

Research Question 4: To what extent do students with significant disabilities independently and reliably access the videos?

Method

Setting

All participants were selected from a segregated school serving secondary school students with significant disabilities in a major metropolitan area in the U.S. Midwest area. Research sessions were conducted in a conference room (15 ft × 9 ft) in the participants’ school. The conference room had one table, two chairs, one desktop, one two-drawer chest, one bookshelf, and two sets of couches.

Participants

Procedures were approved by a university Institutional Review Board, and participants consented to their participation prior to the study. Inclusion criteria required that students (a) were between the ages of 14 and 22, (b) qualified to take the state’s alternate assessment for students with significant cognitive disabilities, (c) had sufficient fine motor skills to operate a tablet computer, and (d) were recommended by their education team as someone who would benefit from VBI.

Three participants met the criteria. All three participated in community-based vocational skill learning experiences at least once a week during the school year. Two participants had no experience with self-directed VP; the third had experience through a research project but did not use this approach as part of his regular school instruction. No systematic VBI was observed in all participants’ classrooms.

Irvin was a 21-year-old Black male student diagnosed with autism and developmental delays. He was reported to have a moderate intellectual disability (Intelligence Quotient = 40) and a low level (i.e., 31) of standard Adaptive Behavior Composite score of the Vineland Adaptive Behavior Scales–Third Edition (Vineland-III; Sparrow et al., 2016). He communicated mainly through his body language (e.g., gesturing, waving, nodding) or the program, Proloquo2Go, installed on an iPad mini. Irvin had self-directed VP learning experiences through previous research projects. His teacher reported that he heavily relied on adult prompting at school.

Avi was a 19-year-old White female student with multiple disabilities. Her Composite Intelligence Index of Reynolds Intellectual Assessment Scales–Second Edition (RIAS-2; Raines et al., 2018) was 40, Verbal Intelligence Index 40, and Nonverbal Intelligence Index 43. All three indexes indicated her severe deficits in overall development. She communicated her needs vocally and could maintain a basic conversation. She had not received systematic VBI before.

Nova was a 20-year-old Black female student with autism, moderate intellectual disability, and seizure disorder. Her Autism Index score of Gilliam Autism Rating Scale–Third Edition was 87. Her Adaptive Behavior Composite score of Vineland-III (Sparrow et al., 2016) was 32 (low = 20–70). Nova communicated her needs mainly through gestures and the program Proloquo2Go, which was installed on an iPad. She had no learning experience with systematic VBI.

Materials and Equipment

Video Prompts and Player

Videos of tasks were recorded from the point-of-view perspective with a Windows 7 Surface tablet/iPhone 13, edited in iMovie software, and played on a Windows 7 Surface. Two sets of videos were made for skill demonstration. The first set of nine video clips was made for the VP intervention for a nine-step badge-making skill (see Table 1 for a detailed task analysis of this skill). Each clip demonstrated one step of the task and ranged from 18 s to 40 s. A second set of videos focused on asking for help in the context of the badge-making task. One individualized video was made for each student. These videos were not broken into individual steps but showed the entire process of badge-making. Each video was about 3 min long and included a scenario in which the student required help and asked for help. Videos were individualized by showing the request for help in a form that was compatible with the student’s communication repertoire. More specifically, the video made for Irvin included how to complete the badge and request help through his own AAC device; the video made for Avi included the badge task and verbally requesting help when there were missing materials; for Nova, the video demonstrated the badge-making process and requesting help from her own AAC device.

Table 1.

Task Analysis for a Badge-Making Task.

Find the right nametag

Put the right nametag into the plastic holder

Attach a lanyard

Find the right nametag

Put the right nametag into the plastic holder

Attach a lanyard

Find the right nametag

Put the right nametag into the plastic holder

Attach a lanyard

Task Materials

Materials for the tablet training included three small round craft containers (two were 1 in × 2 in; one was 1 in × 1.5 in) and six counting cubes (0.4 in × 0.4 in). Materials for the badge-making included three green index cards with the target names (i.e., Alex, Eric, and Cate; 2.5 in × 3.5 in), nine name tags (i.e., Aden, Aaron, Alex, Evan, Eric, Eden, Cate, Cole, and Clay; 2 in × 3 in), three yellow index cards with a sorting letter on each card (i.e., A, E, C; 3 in × 5 in), three green click-type lanyards (3 ft), and three card holders (3 in × 4 in).

Dependent Measure and Data Collection

There were two primary dependent variables. One dependent variable was the independent completion of the badge-making task (i.e., the percentage of task steps completed accurately before the error correction procedure). Correct responding was defined as (a) initiating a step within 5 s of hearing the task direction, completing the prior step, or watching a video clip, and (b) performing the step correctly. Errors were categorized as latency errors (i.e., students failed to initiate the task within 5 s) or topographic errors (i.e., step performed incorrectly). The other dependent variable was requesting help (i.e., verbally expressing “I need help” or clicking the “Help” button on their AAC devices) when there was a missing material (i.e., number of accurate responses among four opportunities). Four items were randomly picked as missing materials in each session.

In addition, we collected several other descriptive variables: (a) the duration of each session except ones in the baseline phase (i.e., how long did it take for a participant to complete a task), (b) students’ tablet operation behavior during intervention phases (i.e., the percentage of correct responses of playing the right video before performing the matching step in a task), and (c) students’ self-initiated self-fading behavior during intervention phases. The self-fading behavior was defined as correct responses with self-initiated reduced VP (i.e., fading VP partially or fading VP completely). When a student correctly performed a step without playing the matching video clip, it was counted as correct responses with completely faded VP; when a student only watched part of a video and correctly performed a step, it was counted as correct responses with partially faded VP.

Experimental Design and Analysis

This study used two multiple-probe designs to focus on two different sets of research questions. The first focused on the efficacy of VP in the acquisition of a badge-making skill, and the second focused on VM in the acquisition of a help-requesting skill. Phases consisted of baseline for the badge-making skill, VP for the badge-making skill, VP for the badge-making skill and baseline for help-requesting skill, VP for the badge-making skill, and VM for help-requesting skill and fading for both badge-making skill and help-requesting skill.

Baseline data were collected for all participants at the same time. When all participants were ready for the intervention (i.e., a minimum of three data points indicating a stable or decreasing trend), we adopted the case randomization (Levin & Kratochwill, 2021) and used an online randomization generator tool (https://www.random.org/sequences/) to identify the sequence for participants to receive the intervention. This randomization procedure enhances the research design by mitigating researcher bias (Kratochwill & Levin, 2010). The mastery criterion/phase change criterion for the following phases was that participants would perform the task with 100% accuracy for four of five sessions: the phase of badge-making with VP, the phase of badge-making with VP and help-requesting with VM, and the fading phase. Sessions were conducted up to 4 times per day, 4 days a week, based on participants’ availability. The data were collected between April 27, 2022, and May 23, 2022.

We conducted visual analysis within each phase (i.e., level, trend, and variability) and across conditions (i.e., the immediacy of effect, overlap, and data consistency; Kratochwill et al., 2013). We adopted the percentage of nonoverlapping data (PND; Scruggs et al., 1987) to demonstrate the effect size and augment the visual analysis. The effectiveness was reported based on the following criteria: effective (i.e., PND > 70), questionable effectiveness (i.e., PND ranging from 50 to 70), and no observed effect (i.e., PND < 50; Scruggs & Mastropieri, 1994). Overall, we identified functional relations between self-directed VP and acquisition of the badge-making task and VM on the help-requesting skill.

Procedures

Identifying Tasks

The vocational task was identified through a collaborative process involving records review, discussions with teachers, and informal observations of students’ behaviors. Ian’s job skills were evaluated through VocFit.com before, which showed that he had strength in positions like library assistant, warehouse worker, or dishwashing worker. Nova’s work preferences were evaluated through the Unique Learning System transition assessment and teacher observations that indicated that she preferred to work in a clean, quiet indoor environment, and she was also interested in watching cooking videos and working on complex puzzles. Avi’s teacher mentioned Avi could be a good candidate for office-related jobs. She was capable of doing tasks like sorting and typing and could maintain simple conversations verbally. After discussing with the participants’ teachers, a novel clerical task (i.e., making name badges) was identified for all participants.

Baseline

During the baseline condition, the researcher started the session with the direction (i.e., “Please make the name badges.”) and thanked participants after signals of completion (i.e., verbally request or sign “all done,” no initiation or stop working for more than 5 s). No intervention was delivered to any students during baseline.

Tablet Training

Tablet operation training was conducted through a nine-step assembly task (i.e., putting specific numbers of cubes into different craft containers). The training consisted of one errorless teaching session and several test sessions until participants met the criterion (i.e., correctly operating the tablet for three consecutive sessions). During the one-trial errorless teaching session, the researcher used the controlling prompt (i.e., a type of prompt that would ensure the accuracy of a response) to prompt participants to operate the tablet and perform the assembly task. During the test sessions, the controlling prompt would be offered if participants made an error when playing the video. All participants achieved this mastery criterion within four test sessions.

Video Prompting for Badge-Making Task

During this phase, the researcher started the session with the direction (i.e., “Please make the name badges.”) and played the first video. Participants were expected to independently play and complete the following steps at a step-by-step pace (i.e., watch one video clip at a time and perform the matching step before proceeding to the next video clip). For the tablet operation behavior, if there was an error (i.e., no initiation within 5 s or clicked a wrong button), the researcher would offer a controlling prompt (i.e., gestural prompt). For the task performance, if a participant failed to initiate the step within 5 s or completed the step inaccurately for the first time of each step, the participant would be prompted to rewatch the specific step and redo it. If the student failed to perform the step correctly for the second time, the researcher would block the participant’s view and finish the step for the participant. By the end of each session, participants would receive explicit verbal praise when they achieved no less than 60% of task completion (e.g., “Great job making the labels!”) or a general comment when the performance remained close to the baseline level (e.g., “Thanks for making the labels!”).

Baseline for Help-Requesting Skill and VP for Badge-Making Task

Once all participants met the mastery criterion for the badge-making skill, they started the second skill learning (i.e., requesting help when there were missing items). This phase was similar to the prior intervention phase except that four help-requesting opportunities (i.e., when an item to complete a step was missing) were randomly inserted into each task. A correct response was defined as a participant requesting help verbally or through an AAC device within 5 s. An error was counted when a participant failed to request help within 5 s of the initial direction or upon completing a prior step. No VM intervention was offered to help-requesting skill.

Video Modeling for Help-Requesting Skill and VP for Badge-Making Task

After all participants achieved a stable or decreasing trend within three consecutive sessions during baseline, the first participant started to receive the intervention. The intervention sequence followed the previous order on badge-making learning. The second set of videos were used for this phase (i.e., one individualized video for each participant demonstrating completing badges and requesting help when there were missing items). In this phase, sessions started with a video demonstration on how to request help (i.e., “How to ask for help?”) and then participants practiced this requesting skill. If the participant failed to respond to it, the researcher would use the controlling prompt to correct the error until they could perform it independently once. After this brief practice, participants would perform the whole badge-making task with four requesting opportunities that were randomly embedded in the task. If there was an error on completing the badges (i.e., no initiating the step within 5 s or completing the step incorrectly), a controlling prompt would be offered; if an error occurred on requesting help (i.e., no requesting help within 5 s of the initial direction or upon the completion of the prior step), the researcher would offer a controlling prompt (i.e., verbally prompting Avi: “You can ask for help”; pointing to the help request button on AAC devices for Irvine and Nova).

Fading

This phase was similar to the baseline. The controlling prompt (i.e., verbally prompting Avi: “You can ask for help”; pointing to the help request button on AAC devices for Irvine and Nova) would be offered if participants’ performance remained low.

Interobserver Agreement and Procedural Integrity

The researcher who was a fourth-year doctorate student in special education collected all primary data in this study. Another special education doctorate student who received data collection training in a similar self-directed VP research project (Sun & Brock, 2024) collected the secondary data for interobserver agreement (IOA) and procedure integrity (PI) data. Therefore, training in this project mainly consisted of verbal instruction. When IOA or PI dropped below 90%, the data collectors would discuss the disagreement and review the data collection protocol until a consensus was reached.

Both IOA and PI data were collected during 32% of all sessions distributed across conditions and participants. The IOA was calculated by dividing the total number of agreements by the sum of total sessions and multiplying by 100 (Ledford & Gast, 2018). The mean IOA was 93%, ranging from 80% to 100%. The PI was calculated by dividing the total number of steps completed correctly by the sum of steps completed correctly plus steps completed incorrectly or not completed and multiplying by 100 (Ledford & Gast, 2018). The mean PI was 100%.

Results

Figure 1 shows the percentage of total correct responses in each session, the duration of each session, and correct responses of requesting help in each session. Besides, participants’ self-fading behavior and tablet operation behavior were graphed in Figure 2. All participants were able to achieve an immediate increase in the level of skill acquisition when the intervention was introduced and maintained a high level of stability in skill performance with few variabilities.

Irvin

Making badges was a novel skill for Irvin, and he could not perform any step during the baseline. However, when the intervention of self-direct VP with error correction was introduced, an immediate increase with small variabilities in the level of badge-making skill was observed (i.e., ranging from 78% to 100%), with no overlapping data points with the initial baseline. Irvin did not request help through his AAC device when there were missing materials before introducing the intervention of VM. With VM, he immediately requested help for 50% of the opportunities. Although there were some variabilities for the first three sessions, he requested help for all opportunities at the fourth session and met the criterion at the seventh intervention session. There was no overlap between baseline and intervention sessions, and the PND was 100% for both badge-making and help-requesting skills. He maintained both skills after the interventions were removed. The average duration of intervention sessions for badge-making only was 5.48 min (ranging from 4.5 to 7.53 min); for both badge-making and help-requesting, 4.66 min (ranging from 3.5 to 7.25 min); and for fading phase, 1.92 min (ranging from 1.33 to 2.47 min). The overall decreasing trend in duration indicated that Irvin achieved fluency with more practice over time (see Figure 1).

Irvin initiated self-fading behavior from the first session in the intervention phase. Self-fading behavior was detected in all intervention sessions. Complete fading behavior was detected in 94% (i.e., 15/16) of intervention sessions ranging from 22% to 77% (see Figure 2). He maintained a high level of performance in operating the tablet with 100% accuracy in 14 of 16 sessions. His average accuracy in operating a video device was 99% (ranging from 89% to 100%). Both errors were topographic errors (i.e., clicked the wrong button) that occurred among the first nine sessions. He maintained 100% operation accuracy in the last seven sessions.

Avi

Avi could not perform any step of making badges correctly during the baseline. However, she immediately achieved a high level of stability (i.e., 100% accuracy) after the self-directed VP intervention was introduced and maintained the skill for all the following sessions without variabilities. During the baseline for help-requesting skills, Avi did not verbally request any help. However, with the VM being introduced, Avi acquired the skill immediately with 100% accuracy through all the following sessions without variabilities. There was no overlap between baseline and intervention sessions and the PND was 100% for both badge-making and help-requesting skills. When both VM and VP were removed in the fading session, Avi still maintained both badge-making skill and help-requesting skills with 100% accuracy. The average duration of intervention sessions for badge-making only was 5 min (ranging from 4.75 to 5.31 min); for both badge-making and help-requesting, 4.79 min (ranging from 4.55 to 5.06 min); and for fading phase, 1.86 min (ranging from 1.63 to 2.16 min). Overall, the duration of each tracked session within the same phase remained stable with a slightly decreasing tendency (see Figure 1).

Avi had initiated partial self-fading behavior (i.e., watching a video while performing the matching step of the task) since the first session of the VP phase. She maintained the partial self-fading behavior for every step throughout the sessions in the same topography except for one session in which 11% (i.e., 1/9) of steps were performed after she fully watched a video clip (see Figure 2). She remained 100% accurate in device operation across the sessions.

Nova

During baseline, Nova was simply sorting green index cards (i.e., visual cues for participants to look for specific name tags) to the name tag piles. Nova could not perform any step of making badges during the baseline. However, there was an immediate increase in the level of skill acquisition and Nova achieved 100% accuracy for the first intervention session and met the criterion in the fifth VP only session. During the baseline for help-requesting skills, Nova did not use her AAC device to request any help. With the VM being introduced, Nova started requesting help for half opportunities for the first VM invention session and maintained 100% accuracy since the second VM session. There was no overlap between baseline and intervention sessions, and the PND was 100% for both badge-making and help-requesting skills. When both VM and VP were removed in the fading session, Avi maintained both badge-making skills and help-requesting skills with 100% accuracy. The average duration of intervention sessions for badge-making only was 5.44 min (ranging from 4.81 to 6.61 min); for both badge-making and help-requesting, 4.94 min (ranging from 4 to 6.5 min); and for fading phase, 1.79 min (ranging from 1.63 to 1.93 min; see Figure 1).

There was a noticeable decreasing trend in the phase combining both VM and VP, which also reflected Nova’s self-initiated fading behavior. She started partially and completely fading the VP while maintaining a high level of skill performance (see Figure 2). Her average accuracy in operating a video device was 96% (range = 77%–100%). All five errors were latency errors that occurred during the first 12 sessions. She maintained 100% operation accuracy in the last four sessions.

Discussion

Both technical knowledge and soft skills are important for employment. However, research has mostly targeted technical skills (e.g., clerical tasks, cleaning tasks) acquisition among students with significant disabilities and has rarely targeted technical and soft skills simultaneously (Sun & Brock, 2023b). This study evaluated both the effects of self-directed VP on technical skills and VM on soft skills. Both interventions were effective, and two of three participants were able to independently and reliably manipulate the video technology. These findings extend the literature regarding the application of self-directed VBI on employment skills in a number of ways.

First, all participants mastered the badge-making skill with the self-directed VP that furthered this intervention toward the evidence-based practice (EBP) on vocational instruction. When combined with preexisting evidence, there are six studies with 17 participants demonstrating the functional relations between self-directed VP and employment skill acquisition among secondary students with significant disabilities (i.e., Bereznak et al., 2012; Heider et al., 2019; Sun & Brock, 2024; Van Laarhoven et al., 2009; Yakubova et al., 2019). Based on Council for Exceptional Children (CEC) criteria for EBP requiring five single-case design studies showing positive effects across 20 participants (Cook et al., 2014), one or more single-case studies with positive effects across at least three more participants are yet to be conducted to establish the EBP status of self-directed VP on the vocational instruction among secondary students with significant disabilities.

Second, all participants mastered the skill of requesting help with the VM intervention. Prior studies have demonstrated the effectiveness of VM on social skills such as greetings, using service phases, and responding to feedback or clarifying unclear instructions (Bross et al., 2020; Park et al., 2020). Differing from other studies, this study did not test the generalization effect separately. Instead, it combined VM and the loose training generalization method during the intervention phase. During training, four help-requesting opportunities were randomly embedded in this nine-step task that did not match the order presented in the videos. Using the same method to request help across different stimuli may simplify the generalization process and promote success. This aligns with the previous finding that the generalization tended to occur during stimulus generalization scenarios (Park et al., 2020). However, due to the limited evidence on employment-related social skill acquisition (Sun & Brock, 2023b), it is hard to draw strong conclusions about the most effective path for skill generalization. In the future, researchers may further explore factors and avenues that may contribute to generalization success in employment-related social skills.

This is one of the first studies targeting both technical and job-related social skills of a job task with the combination of VP and VM. Most studies have focused either on technique skills preparation or on “soft” skills training (Sun & Brock, 2023a; Yakubova & Taber-Doughty, 2017). While these findings are critical to advance vocational training for individuals with significant disabilities, discovering venues for equipping them with both “hard” skills and “soft” skills may facilitate their CIE (Wehman et al., 2018). The inclusion of soft skills instruction may be especially important, as employers often report that it is soft skills—and not technical skills—that often limit the success of employees (Robles, 2012). This combined approach may serve as a model for how to provide more comprehensive instruction on employment skills.

Third, all participants have varied levels of self-fading behavior (i.e., self-initiated, completely skipping the video or partially fading the video). One participant had complete fading behavior across most of the sessions. One participant only partially faded prompt across all sessions. The other participant started with partial fading and skipped some videos in the last four sessions. These findings align with those in previous studies that participants tend to fade the prompt to a less intrusive level as they progressively acquired the skill (Mechling et al., 2009; Van Laarhoven et al., 2007). In addition, the current study systematically captured each individual’s self-fading pace and magnitude. This novel evidence further demonstrates that self-direct VP has the unique advantage of meeting individual learner’s need and contributing to self-initiated learning among individuals with significant disabilities.

Fourth, the mixed results on the device operation add another important set of evidence in the current literature. Two of three participants operated the device reliably and fluently, while the other participant (i.e., Nova) exhibited latency errors on device operation in 31% of sessions. Nova’s teacher reported that Nova often sought reassurance and was overly dependent on instructors’ prompts during instruction. This habit may have led to latency errors. So far, the literature has not consistently reported the accuracy of device operation in self-directed VP sessions. Some studies reported complete success or mixed results on device operation after training (Cullen et al., 2017; Heider et al., 2019), while others only reported training results (Cannella-Malone et al., 2013; Sun & Brock, 2023a). The strategies mostly used in device training were prompting and error correction. Also, trainers tended to use an easy or familiar task during the device training to reduce the acquisition demand. Future studies may collect more evidence on the acquisition and maintenance of device training, which can improve the reliability of individuals’ device operation.

Limitations and Future Directions

There are two significant limitations to this study. First, there was no social validity data directly collected from participants or their instructors. While conducting a survey among participants may be the most straightforward method to have their opinions, it may not be an effective way for participants in this study due to their limited critical thinking skills and tendency to express “yes” to instructors’ directions. However, given the features of the self-directed VBI that requires self-initiative and self-regulation, it is important to collect reliable data on how individuals perceive this intervention. This may further contribute to generalizing self-directed intervention across settings and skills. Therefore, to collect data better reflecting participants’ opinions, future studies may focus on two directions: exploring metrics for measurement and improving the critical thinking skills of participants (e.g., teaching them how to express disagreement or rejection) with significant disabilities. Second, due to the limited study time, no separate generalization phase was conducted. Help-requesting skill is a basic but very important skill across settings. However, the performance level may also be impeded by an individual’s personality, a setting change, or a team personnel change. Therefore, it is important to have participants practice this skill across settings with different support personnel. In addition, due to the feature of social skills, there will be expression variations across scenarios. Therefore, exploring effective avenues to improve response generalization is worth noting in future research.

Implications for Practice

These findings have two major implications for practitioners and caregivers who are helping with vocational skill training among young adults with significant disabilities. First, self-directed VP can be applied to a multistep task acquisition and to boost students’ learning independence. For example, teachers can train students who are overdependent on the adult prompt to use the device to deliver the prompt themselves. In addition, with easy access to video players nowadays, self-directed VP can be used during school training and generalized to a job site. This will contribute to the generalization success and resolve the potential low staff support issue at the job site. Video modeling can be used for social skill training. Different from technical skills, social interaction involves more variations. Mastering a social skill demands both stimulus and response generalization. Using VM for multiple exemplar demonstrations can help clarify and model the learning expectation. Second, young adults with significant disabilities have great potential for self-initiated learning. Offering different levels of support and allowing students to choose their prompt level may promote their self-determination and self-initiation ability. When learners perform a task with 100% accuracy and start self-fading the prompt, trainers may insert trials without intervention to avoid over prompt.

Conclusion

Lack of employment skills training is one of the key issues that hinder young adults with significant disabilities from job opportunities (Adams et al., 2019). This study adds a novel set of evidence on using VBI in employment skill training by systematically tracking self-fading behavior, device operation performance, and combining VM and the loose training method. Researchers should continue to explore effective training methods and combine more technology components into VBI to promote the employment outcomes of young adults with significant disabilities.

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.

ORCID iD

Xiaoning Sun

References

Adams

Corbin

O’Hara

Park

Sheppard-Jones

Butler

Umeasiegbu

McDaniels

Bishop

M. L.

(2019). A qualitative analysis of the employment needs and barriers of individuals with intellectual and developmental disabilities in rural areas. Journal of Applied Rehabilitation Counseling, 50(3), 227–240. https://doi.org/10.1891/0047-2220.50.3.227

Anderson

Butt

Sarsony

(2021). Young adults on the autism spectrum and early employment-related experiences: Aspirations and obstacles. Journal of Autism and Developmental Disorders, 51(1), 88–105. https://doi.org/10.1007/s10803-020-04513-4

Barczak

M. A.

Cannella-Malone

H. I.

(2022). Self-management of vocational skills for people with significant intellectual disabilities: A systematic review. Journal of Intellectual Disabilities, 26(2), 470–490. https://doi.org/10.1177/1744629520987768

Bereznak

Ayres

K. M.

Mechling

L. C.

Alexander

J. L.

(2012). Video self-prompting and mobile technology to increase daily living and vocational independence for students with autism spectrum disorders. Journal of Developmental and Physical Disabilities, 24(3), 269–285. https://doi.org/10.1007/s10882-012-9270-8

Biard

Cojean

Jamet

(2018). Effects of segmentation and pacing on procedural learning by video. Computers in Human Behavior, 89, 411–417. https://doi.org/10.1016/j.chb.2017.12.002

Blustein

D. L.

(2008). The role of work in psychological health and well-being: A conceptual, historical, and public policy perspective. American Psychologist, 63(4), 228–240. https://doi.org/10.1037/0003-066X.63.4.228

Bross

L. A.

Travers

J. C.

Wills

H. P.

Huffman

J. M.

Watson

E. K.

Morningstar

M. E.

Boyd

B. A.

(2020). Effects of video modeling for young adults with autism in community employment settings. Career Development and Transition for Exceptional Individuals, 43(4), 209–225. https://doi.org/10.1177/2165143420941488

Browder

D. M.

Shapiro

E. S.

(1985). Applications of self-management to individuals with severe handicaps: A review. Journal of the Association for Persons with Severe Handicaps, 10(4), 200–208. http://doi.org/10.1177/154079698501000403

Buttler

(2022). Employment status and well-being among young individuals. Why do we observe cross-country differences? Social Indicators Research, 164, 409–437. https://doi.org/10.1007/s11205-022-02953-2

10.

Cannella-Malone

H. I.

Brooks

D. G.

Tullis

C. A.

(2013). Using self-directed video prompting to teach students with intellectual disabilities. Journal of Behavioral Education, 22(3), 169–189. https://doi.org/10.1007/s10864-013-9175-3

11.

Cannella-Malone

H. I.

Fleming

Chung

Y.-C.

Wheeler

G. M.

Basbagill

A. R.

Singh

A. H.

(2011). Teaching daily living skills to seven individuals with severe intellectual disabilities: A comparison of video prompting to video modeling. Journal of Positive Behavior Interventions, 13(3), 144–153. https://doi.org/10.1177/1098300710366593

12.

Cannella-Malone

H. I.

Sigafoos

O’Reilly

de la Cruz

Edrisinha

Lancioni

G. E.

(2006). Comparing video prompting to video modeling for teaching daily living skills to six adults with developmental disabilities. Education and Training in Developmental Disabilities, 41(4), 344–356. https://www.jstor.org/stable/23879661

13.

Card

Moran

Newal

(1983). The psychology of human computer interaction. Erl-baum. https://doi.org/10.1201/9780203736166

14.

Cihak

D. F.

Kessler

K. B.

Alberto

P. A.

(2007). Generalized use of a handheld prompting system. Research in Developmental Disabilities, 28(4), 397–408. https://doi.org/10.1016/j.ridd.2006.05.003

15.

Cook

B. G.

Buysse

Klingner

Landrum

McWilliam

Tankersley

Test

(2014). Council for Exceptional Children: Standards for evidence-based practices in special education. Exceptional Children, 80(4), 504–511. https://doi.org/10.1177/0014402914531388

16.

Cullen

J. M.

Alber-Morgan

S. R.

Simmons-Reed

E. A.

Izzo

M. V.

(2017). Effects of self-directed video prompting using iPads on the vocational task completion of young adults with intellectual and developmental disabilities. Journal of Vocational Rehabilitation, 46(3), 361–375. https://doi.org/10.3233/JVR-170873

17.

Gilson

C. B.

Carter

E. W.

Biggs

E. E.

(2017). Systematic review of instructional methods to teach employment skills to secondary students with intellectual and developmental disabilities. Research and Practice for Persons with Severe Disabilities, 42(2), 89–107. https://doi.org/10.1177/1540796917698831

18.

Heider

A. E.

Cannella-Malone

H. I.

Andzik

N. R.

(2019). Effects of self-directed video prompting on vocational task acquisition. Career Development and Transition for Exceptional Individuals, 42(2), 87–98. https://doi.org/10.1177/2165143417752901

19.

Hughes

M. A.

Alberto

P. A.

Fredrick

L. L.

(2006). Self-operated auditory prompting systems as a function-based intervention in public community settings. Journal of Positive Behavior Interventions, 8(4), 230–243. https://doi.org/10.1177/10983007060080040501

20.

Kellems

R. O.

Cacciatore

Hansen

B. D.

Sabey

C. V.

Bussey

H. C.

Morris

J. R.

(2021). Effectiveness of video prompting delivered via augmented reality for teaching transition-related math skills to adults with intellectual disabilities. Journal of Special Education Technology, 36(4), 258–270. https://doi.org/10.1177/0162643420916879

21.

Kratochwill

T. R.

Hitchcock

J. H.

Horner

R. H.

Levin

J. R.

Odom

S. L.

Rindskopf

D. M.

Shadish

W. R.

(2013). Single-case intervention research design standards. Remedial and Special Education, 34(1), 26–38. https://doi.org/10.1177/0741932512452794

22.

Kratochwill

T. R.

Levin

J. R.

(2010). Enhancing the scientific credibility of single-case intervention research: Randomization to the rescue. Psychological Methods, 15(2), 124–144. https://doi.org/10.1037/a0017736

23.

Ledford

J. R.

Gast

D. L.

(2018). Single case research methodology: Applications in special education and behavioral sciences (3rd ed.). Routledge.

24.

Levin

J. R.

Kratochwill

T. R.

(2021). Randomized single-case intervention designs and analyses for health sciences researchers: A versatile clinical trials companion. Therapeutic Innovation & Regulatory Science, 55(4), 755–764. https://doi.org/10.1007/s43441-021-00274-z

25.

Mechling

L. C.

Ayres

Bryant

Foster

(2014). Comparison of the effects of continuous video modeling, video prompting, and video modeling on task completion by young adults with moderate intellectual disability. Education and Training in Autism and Developmental Disabilities, 49, 491–504. http://www.jstor.org/stable/24582346

26.

Mechling

L. C.

Gast

D. L.

Seid

N. H.

(2009). Using a personal digital assistant to increase independent task completion by students with autism spectrum disorder. Journal of Autism and Developmental Disorders, 39(10), 1420–1434. https://doi.org/10.1007/s10803-009-0761-0

27.

Moss

Tilly

(1996). “Soft” skills and race: An investigation of black men’s employment problems. Work and Occupations, 23(3), 252–276. https://doi.org/10.1177/0730888496023003002

28.

National Core Indicators. (2022). National Core Indicators® 2020–21 In-Person Survey (IPS) Work. https://www.nationalcoreindicators.org/upload/core-indicators/IPS_2020-21__4_Employment.pdf

29.

Nord

Luecking

Mank

Kiernan

Wray

(2013). The state of the science of employment and economic self-sufficiency for people with intellectual and developmental disabilities. Intellectual and Developmental Disabilities, 51(5), 376–384. https://doi.org/10.1352/1934-9556-51.5.376

30.

Park

Bouck

E. C.

Duenas

(2020). Using video modeling to teach social skills for employment to youth with intellectual disability. Career Development and Transition for Exceptional Individuals, 43(1), 40–52. https://doi.org/10.1177/2165143418810671

31.

Payne

Cannella-Malone

H. I.

Tullis

C. A.

Sabielny

L. M.

(2012). The effects of self-directed video prompting with two students with intellectual and developmental disabilities. Journal of Developmental and Physical Disabilities, 24(6), 617–634. https://doi.org/10.1007/s10882-012-9293-1

32.

Raines

T. C.

Reynolds

C. R.

Kamphaus

R. W.

(2018). The reynolds intellectual assessment scales, Second Edition, and the reynolds intellectual screening test, Second Edition. In Flanagan

D. P.

McDonough

E. M.

(Eds.), Contemporary intellectual assessment: Theories, tests, and issues (4th ed., pp. 533–552). The Guilford Press.

33.

Robles

M. M.

(2012). Executive perceptions of the top 10 soft skills needed in today’s workplace. Business Communication Quarterly, 75(4), 453–465. https://doi.org/:10.1177/1080569912460400

34.

Scruggs

T. E.

Mastropieri

M. A.

(1994). The utility of the PND statistic: A reply to Allison and Gorman. Behaviour Research and Therapy, 32(8), 879–883. https://doi.org/10.1016/0005-7967(94)90169-4

35.

Scruggs

T. E.

Mastropieri

M. A.

Casto

(1987). The quantitative synthesis of single-subject research: Methodology and validation. Remedial and Special Education, 8(2), 24–33. https://doi.org/10.1177/074193258700800206

36.

Shepard

Migliore

Butterworth

Winsor

(2020). Employment and economic outcomes by race, ethnicity, and gender for individuals with and without disabilities, 2017. University of Massachusetts Boston, Institute for Community Inclusion. https://www.thinkwork.org/sites/default/files/files/trends_2017_F2b.pdf

37.

Sparrow

S. S.

Cicchetti

D. V.

Saulnier

C. A.

(2016). Vineland Adaptive Behavior Scales (3rd ed.). Pearson.

38.

Storey

(2007). Review of research on self-management interventions in supported employment settings for employees with disabilities. Career Development for Exceptional Individuals, 30(1), 27–34. https://doi.org/10.1177/08857288070300010301

39.

Sun

Brock

M. E.

(2023a). Practitioner-implemented video prompting on vocational skills of students with significant intellectual and developmental disabilities. Career Development and Transition for Exceptional Individuals, 0(0). https://doi.org/10.1177/21651434231170540

40.

Sun

Brock

M. E.

(2023b). Systematic review of video-based instruction to teach employment skills to secondary students with intellectual and developmental disabilities. Journal of Special Education Technology, 38(3), 288–300. https://doi.org/10.1177/01626434221094793

41.

Sun

Brock

M. E.

(2024). Young adults with intellectual and developmental disabilities self-delivered video prompts for vocational skill learning [Manuscript submitted for publication]. Department of Educational Studies, The Ohio State University.

42.

Van Laarhoven

Johnson

Van Laarhoven-Myers

Grider

. (2009). The Effectiveness of using a video iPod as a prompting device in employment settings. Journal of Behavioral Education, 18(2), 119–141. https://doi.org/10.1007/s10864-009-9077-6

43.

Van Laarhoven

Van Laarhoven-Myers

Zurita

L. M

. (2007). The effectiveness of using a pocket PC as a video modeling and feedback device for individuals with developmental disabilities in vocational settings. Assistive Technology Outcomes and Benefits, 4(1), 28–45. https://files.eric.ed.gov/fulltext/EJ899364.pdf

44.

Wehman

Taylor

Brooke

Avellone

Whittenburg

Ham

Brooke

A. M.

Carr

(2018). Toward competitive employment for persons with intellectual and developmental disabilities: What progress have we made and where do we need to go. Research and Practice for Persons with Severe Disabilities, 43(3), 131–144. https://doi.org/10.1177/1540796918777730

45.

Yakubova

Leibowitz

Baer

B. L.

Halawani

Lestremau

(2019). Self-directed video prompting and least-to-most prompting: Examining ways of increasing vocational skill acquisition among students with autism spectrum disorder and intellectual disability. Advances in Neurodevelopmental Disorders, 3(3), 246–258. https://doi.org/10.1007/s41252-019-0097-5

46.

Yakubova

Taber-Doughty

(2017). Improving problem-solving performance of students with autism spectrum disorders. Focus on Autism and Other Developmental Disabilities, 32(1), 3–17. https://doi.org/10.1177/1088357615587506