Effectiveness of Video Prompting Delivered via Augmented Reality for Teaching Transition-Related Math Skills to Adults With Intellectual Disabilities

Abstract

The purpose of this study was to determine the effectiveness of a video-based instruction packet for teaching math-based vocational skills delivered through augmented reality (AR) to three adults with intellectual disabilities. The dependent variable was the percentage of steps performed correctly to solve each selected type of math problem. The independent variable was the video-based math intervention delivered via AR, which modeled the individual steps for solving three different multistep math problems: (1) adjusting a recipe to serve a different number of people, (2) calculating salary, and (3) calculating unit prices. Visual and statistical analyses demonstrated a functional relationship between the video-based math intervention and an increase in the percentage of steps completed correctly for each type of problem. All three participants showed significant gains immediately after receiving the intervention and maintained the learned skills following withdrawal of the intervention. Implications for practitioners and further research are discussed.

Keywords

augmented reality video modeling transition math intellectual disabilities

As individuals with disabilities are integrated into society, understanding their challenges and finding evidence-based practices for teaching them daily living skills have become essential. Basic numeracy is a skill area considered necessary for employment as well as an essential component of independent living (Wei et al., 2013), one of the most common necessities for individuals with or without disabilities. Over the past few years, emphasis has increased on teaching math skills to adults with disabilities, and more research is needed to define effective methods to deliver math instruction to this population.

Math Challenges for Individuals With Disabilities

According to the revised (2014) Diagnostic and Statistical Manual of Mental Disorders, fifth edition, as well as the American Association on Intellectual and Developmental Disabilities, intellectual disability (ID) is characterized by significant limitations and global deficits in both intellectual and adaptive functioning (Schalock et al., 2010). The majority of individuals diagnosed with ID often struggle in understanding and correctly performing math skills (King et al., 2016). Both the nature of current mathematics curriculum utilized by most schools and the nature of their disability contribute to these difficulties.

According to Patton et al. (1997), the math curriculum presented in schools, specifically at the secondary level, is primarily designed for students who will continue their studies by going to college. Such curriculum ignores some of the independent living skills individuals need to function at home, in the community, and on the job. For many students without disabilities, such skills are learned in informal ways including deduction from other more complex skills. However, students with disabilities who are not specifically and directly taught skills pertinent to independent living will lack the knowledge necessary to implement them (Patton et al., 1997). Patton and colleagues (1997) further advocated that math skills pertinent to home, community, and workplace must be taught as a matter of routine to all students with disabilities. Too much time and effort are spent in trying to help students with disabilities catch up with the core curriculum and too little in focusing instruction on math skills that will benefit them in practical ways.

According to the National Center for Education Statistics, during the 2017–2018 school year, 6% of the students (ages 3–21) enrolled in special education in the United States had a diagnosis of ID and were served under the Individuals with Disabilities Education Act (Kena et al., 2016). As shown by this statistic, the percentage of individuals with a diagnosis of ID is significant, necessitating attention to their levels of performance and research into ways to improve it. The No Child Left Behind Act (2002) is pressuring the public school system to help all students, in particular those with disabilities, reach proficiency levels in math and in other subjects as well (Wei et al., 2013). Research also shows that students with disabilities exhibit a slower mathematics achievement growth than students without disabilities (Schulte & Stevens, 2015). The National Assessment of Educational Progress conducted by the U.S. Department of Education in 2015, also known as the Nation’s Report Card, documented that 92% of 8th grade students and 94% of 12th grade students with disabilities were below proficient levels in mathematics. When compared to 67% of 8th grade students and the 75% of 12th grade students without disabilities performing below proficiency, the significantly poorer performance of the students with disabilities demonstrates the importance of change.

According to King et al. (2016), a possible explanation for the limited percentage of proficient performance in mathematics of students with and without disabilities is the increase of expectations expressed in the recently adopted Common Core State Standards in mathematics. Due to the increasing gap between expectation and performance, there is a rising need for targeted remediation and for development of new instructional approaches. Another possible explanation for the limited percentage of proficient mathematics performance in students with disabilities is the abstract nature of mathematics and the tendency of individuals with disabilities to rely on concrete experience. In addition to the difficulty they have with abstractions, students with disabilities must rely on multiple prerequisite skills simultaneously, despite deficits many have in the areas of semantics and working memory (Yakubova et al., 2015).

Video-Based Interventions for Individuals With ID

Many individuals with disabilities, including those with ID, struggle in forming social relations (Walton & Ingersoll, 2013), which can affect ways they attend to instructional methods which are based on interaction between the instructor and the student. Both direct instruction and in vivo modeling require interaction between the instructor and the learner, and difficulty attending to these prevalent methods without the opportunity to easily watch the instruction presented multiple times leads to limited learning and thus to below proficiency performance (Charlop-Christy et al., 2000; Corbett & Abdullah, 2005). Video-based interventions, such as video modeling (VM) and video prompting (VP), are considered more effective learning tools for this population because they combine observational learning with the apparent tendency of individuals with ID to be particularly responsive to visually cued instruction (Bellini & Akullian, 2007).

Walton and Ingersoll (2013) pointed out that because individuals with disabilities tend to benefit from visually cued instruction and to be less responsive to in vivo behaviors, VM can be effective because it draws on what is known from the social learning literature, as well as on the visual strengths and preferences of these individuals, to teach a variety of social, educational, adaptive, and vocational tasks to them. Video-based instruction also allows students to progress at their own pace and to have opportunities for repeated practice without falling behind their classmates. Furthermore, video-based instruction exposes students to relevant information and limits extra stimuli that may distract them or interfere with their learning. This aspect of video-based instruction is particularly relevant to students with ID who tend to be overly sensitive to the extra stimuli often present with in vivo situations (Yakubova et al., 2015).

According to the available literature, video-based instruction delivered through devices such as iPods and iPads is an effective way of teaching individuals with disabilities. Video-based interventions eliminate some of the problems individuals with ID experience with direct instruction and in vivo modeling, problems caused both by their social interaction difficulties and by the likelihood of distraction by the many stimuli present during the more traditional instructional methods. Learners tend to see the use of technology as reinforcing and thus are overall more responsive to the material presented.

A systematic review of 15 studies addressing the effectiveness of technology devices such as the iPod, iPod Nano, iPod Touch, iPad, and iPhone for delivering instruction to individuals with ID shows that this population can potentially benefit from integrating these devices as instructional and learning tools, thereby enhancing academic, communication, leisure, employment, and transitioning skills. Additionally, using the mobile devices is less stigmatizing than using other forms of assistive technology, a crucial consideration for integrating individuals with disabilities into the community (Kagohara et al., 2013).

VM

Over the past few decades, substantial research has validated VM as an evidence-based practice for teaching a variety of skills to students with disabilities (Bellini & Akullian, 2007; Kellems & Morningstar, 2012). VM involves filming one or multiple individuals (i.e., models)—who can be peers, educators, family members, or the learners themselves (Kellems et al., 2016)—correctly performing a task. The resulting video is used in instructing individuals to perform the task through processes of memorization, imitation, and generalization (Kellems & Morningstar, 2012). VM shows the task performed in its entirety, with the whole sequence of steps shown at once, allowing the learners to get a sense of integration of all the steps in completing the task. After the delivery of the VM intervention, students are asked to perform the task as they have seen it performed in the video (Cannella-Malone et al., 2006).

The effectiveness of VM is often compared to that of in vivo modeling, for which a live model performs a task for the students in the same way they will be asked to perform it. Studies comparing the effectiveness of in vivo modeling versus VM show that VM is more effective for acquisition and generalization of tasks and behaviors and is also more time efficient since teachers can maximize their availability to the classroom as students learn independently (Kellems et al., 2016). Another research-supported advantage of VM compared to in vivo modeling is that prompting and positive reinforcement may not be needed with VM (Allen et al., 2010), as the video instruction is more stimulating with its learning format different from the usual class procedures (Gardner & Wolfe, 2013).

VP

As a form of VM, VP shares most of the features of VM (models performing tasks that the learners will then perform themselves). VP differs by presenting each individual step of a task separately, one clip at a time. With VP, learners have the chance to watch one step and then immediately perform that step before moving on to the next (Kellems et al., 2016; Taber-Doughty et al., 2011). Each clip can be shown multiple times, and learners can advance toward completing a task at their own pace until they become proficient (Banda et al., 2011). Learners with ID often struggle with attention and memory; thus, since a VM needs to be attended to in its entirety, VP with its shorter segments has been determined more effective in supporting the learning process and independence of individuals with such difficulties (Cannella-Malone et al., 2011; Cannella-Malone et al., 2006; Gardner & Wolfe, 2013; Taber-Doughty et al., 2011).

Augmented Reality (AR)

AR is a technology that enhances experience by bringing digital content together with real-world situations. Information about the user’s immediate environment is superimposed with cameras on mobile devices, and information is added from digital sources such as videos and audio (Cihak et al., 2016; Kellems et al., 2020; McMahon et al., 2015; Sommerauer & Muller, 2014). The AR system is characterized by (a) combining the real and virtual worlds, (b) providing interaction in real time, and (c) aligning real objects or places with digital information in 3D (Sommerauer & Muller, 2014).

AR can function as an assistive technology and/or instructional technology, thus supporting students with disabilities in their learning process (Kellems et al., 2019, 2020), in accord with the principles of the universal design for learning. This technology has also shown promise in teaching new academic skills to students with ID in an authentic manner by creating a mobile learning environment that moves with the learner through mobile devices (McMahon et al., 2016). Functions of AR can be utilized to reduce students’ dependence on teachers for help and also increase their independent learning—a critical skill when preparing for life after school (Lin et al., 2016). Wu et al. (2013) asserted that AR has high relevance and that its most promising application is in the field of education through its ability to connect multiple types of technologies and learning pedagogies. Wu et al. also advocates that AR should be viewed as a concept as opposed to a specific type of technology.

Sommerauer and Muller (2014) conducted a large field experiment to test the effects of AR on learning performance and assess its potential for helping students learn formal content in both formal and informal environments. Findings from their research provide evidence of the significant positive effects. AR demonstrates on knowledge acquisition and performance.

In regard to instructing individuals with disabilities, Kellems et al. (2020) conducted a study that utilized an AR-based intervention to teach seven middle school student grade-level math skills. Results indicated a functional relation between the AR-based intervention and the percentage of steps completed correctly for all seven of the participants. In another study, McMahon et al. (2016) tested the functionality of using AR in teaching science vocabulary to postsecondary education students with ID utilizing a single-subject design. All the participants in this study improved in their ability to define and label science terms after receiving the AR intervention, verifying the effectiveness of its use with this population. Also relevant to this study is evidence of AR effectiveness in teaching chained tasks. A chained task is a skill or task that must be performed in a specific order to get the desired outcome. Cihak et al.’s (2016) study examining the effects of using AR in teaching chained tasks to elementary students with disabilities found that the students’ independent performance improved immediately after introduction of the AR intervention with 98% nonoverlapping data, which denotes a high effectiveness for the intervention. Findings of similar studies also suggest that AR is an effective intervention to teach chain tasks to individuals with disabilities.

One strength of AR technology is its ability to act as a platform for delivering instruction or for intervention implementation. Using AR as a platform provides teachers and interventionists the ability to customize the order and rate of content delivery for the desired audience. This study utilized AR as a platform for delivering the instructional videos. The purpose of this study was to explore effectiveness of an intervention comprised of VP delivered via AR.

Research Questions

This study was guided by three research questions:

When teaching multistep math problems to adults with ID, is there a functional relationship between the percentage of steps completed correctly and a VP-based intervention package delivered via AR (HP-Reveal)?

When the intervention is no longer available, are the acquired math skills maintained over time?

According to the participants, how socially valid is the use of a VP intervention delivered via AR for teaching transition-related math skills to adults with ID?

Method

Participants

The participants for this study were three female individuals with ID between the ages of 21 and 24. As reported in Table 1, these participants had full-scale intelligence quotient standard scores of 42 or lower according to the Weschler Intelligence Scale for Children-IV (WISC-IV) as well as low functioning scores in both mathematical achievement and adaptive behaviors. As also reported in Table 1, all three participants had a diagnosis of Down syndrome with a classification of ID. Participants were selected based on their participation in a postsecondary education program for individuals with disabilities and their difficulty with mathematics. None of the participants exhibited problem behaviors that would have affected the conducting of the study.

Table 1.

Participant Information.

Participant	Age	Gender	Classification Category (IDEA)	Cognitive		Achievement (Math)		Adaptive
Participant	Age	Gender	Classification Category (IDEA)	Test	FSIQ	Test	Score	Test	Score
Sophie	24	F	ID	WISC-IV	42	WJ III	44	BASC	39
Ella	21	F	ID	WISC-IV	40	WJ III	39	Vineland	62
Jane	21	F	ID	WISC-IV	42	WJ III	39	Vineland	63

Note. BASC = Behavior Assessment System for Children; FSIQ= Full scale Intelligence Quotient; WJ III- Woodcock Johnson Tests of Achievement; IDEA = Individuals with Disabilities Education Act.

Setting

The participants were enrolled in a postsecondary education program at a large western university. As part of program, the three lived together with other roommates in an off-campus apartment. The study was conducted as an after-school program independent from school curriculum, activities, and credits. The study took place in the dining area of the participant’s residence. The participants worked with the researcher individually, while the other participants and other roommates remained outside the kitchen and dining area as much as possible in order to limit distractions and excessive stimuli. The sessions occurred in the afternoon hours between 3:00 and 6:00 p.m.

Target Tasks

The tasks (dependent variables) were based on the Common Core mathematics standards for sixth and seventh grade. The standards listed below were chosen because of their potential to become functional academic and vocational skills for the participants.

CCSS.MATH.CONTENT.7.RP.A.2.B: “Identify the constant of proportionality (unit rate) in tables, graphs, equations, diagrams, and verbal descriptions of proportional relationships.” Participants were to demonstrate this standard by calculating the unit price of an item.

CCSS.MATH.CONTENT.6.RP.A.3.D: “Use ratio reasoning to convert measurement units; manipulate and transform units appropriately when multiplying or dividing quantities.” Participants were to demonstrate this standard by adjusting the quantities of items needed for a recipe based on variation in number of servings.

CCSS.MATH.CONTENT.6.RP.A.3.D: “Use ratio reasoning to convert measurement units; manipulate and transform units appropriately when multiplying or dividing quantities.” Participants were to demonstrate this standard by calculating daily, weekly, and monthly salary given the rate per hour.

A task analysis was developed for each targeted task, outlining every step necessary for completion. A second researcher or assistant with no involvement in creating the task analyses was then asked to perform the tasks by strictly following the analyses to verify their accuracy. The sequence in which the participants received treatment for the different tasks was randomized.

Research Design

The design selected for this study, a multiple probe design across tasks (Kennedy, 2005), adhered to the standards set by the What Works Clearinghouse for this specific design (Kratochwill et al., 2010). The multiple probe single-case research design is a scientifically valid research method that allows for a systematic introduction of the intervention across multiple tiers of data (e.g., skills or behaviors). This design was selected instead of the multiple baseline design to reduce potential threats to internal validity, specifically testing threats (Gast et al., 2018). Each participant received the intervention in all the selected tasks, but the sequence of the tasks was randomized. Intervention for the second and third tasks started only when the participant had shown proficiency (minimum of 80% accuracy) in the previous task(s). The study also included a maintenance phase to determine if task acquisition would be maintained in the absence of the intervention.

Intervention

The intervention packet consisted of VP and explaining the different steps of the targeted skills to the participants, along with checklists that visually outlined such steps. While the checklists were provided in hard copy, the videos were delivered via iPads through an AR app called HP-Reveal, which provides a platform for on-demand learning. The participants used both the checklist and the videos to complete the targeted tasks step-by-step. The AR app was not the intervention but the means for delivering it. Approximately, 8 weeks elapsed between the first baseline data point to the last maintenance data point. An average of three sessions occurred each week. Each session lasted an average of 30 min in duration.

Materials

Videos were created for each individual step or group of steps that logically fit together. The task analysis served as a script, and the model was filmed performing the tasks exactly as the participants were expected to perform them. A Nikon-One camera on a tripod was used to record the videos. The model in the video was a female who was similar in age and familiar to the participants. The videos were then edited through iMovie® (Version 2.0) on the computer, and voice-over instructions were added to clarify each step. The videos were then organized as a series of segments representing multiple steps (VP). The average length of each video segment was 25 s. It took participants approximately 3–5 min to watch all of the video segments for a given task.

The segments were uploaded to the HP-Reveal app (v 7.0.0.1570036) on an iPad Air Tablet, 16 GB and individually connected to trigger images specifically created for this project. Selected Google images were combined and altered on Microsoft word and Paint to create unique ones. To activate a video, the participant had to slide the iPad screen to unlock it, open the HP-Reveal app, and hover the iPad over the corresponding trigger image. The video would then appear and could be double tapped to go full screen.

The images necessary to trigger the videos through the HP-Reveal app were organized in sequence in booklets. Each segment’s trigger image was associated with a number and color coded to match a checklist that outlined the steps of the task analysis. Thus, the participants were able to choose between watching all the segments in succession pausing in between them to perform that step or series of steps or watching just the segments they felt they needed to perform the task correctly. The checklist served as their guide to determine which trigger image and consequent video were associated with each step.

The researchers created 20 different problem cards for each targeted skill, which were written on paper and individually cut so that participants were able to write on them. Other materials needed were a calculator and writing utensils.

Procedures

Data collection sheets were created based on the format of the task analysis. A pre- and posttest were created: equivalent in difficulty (verified by the outside content expert) and addressing all the targeted skills. The pretest was administered prior to the beginning of the baseline phase as part of the selection process with the intent that those who scored below 50% were eligible for the study. After all scored 0% on the pretest, the three chosen participants were assigned a list of the targeted skills with their respective problems in a randomly assigned order. The posttest was administered after the last maintenance session.

Baseline

During the baseline phase, the participants had access only to the problem cards, writing utensils, and calculator. None of the intervention materials were made available to them at this stage. Participants were presented with the one opportunity to solve each mathematics problem for each skill, one skill at a time, and tasked to solve them. Baseline data on the participants’ performance of all the tasks were collected either for a minimum of six data points, until a clear pattern emerged, or until the data stabilized. Data were considered stable if at least three consecutive data points were consistent in level. Each skill was assessed one time during each baseline session.

Pretraining

After collection of baseline data, participants received direct instruction on how to use the iPad and the specific app (HP-Reveal), then performed a sample test—writing their name by watching videos through the HP-Reveal app—to show their level of proficiency. All demonstrated 100% proficiency in operating the iPad, following the checklist, locating the correct trigger images, and triggering the videos. (This level of accuracy on the overall task was not required.)

Intervention

During the intervention phase, each participant was given an iPad with the HP-Reveal app downloaded and signed into and all the instructional videos stored. The participants independently unlocked the iPad, located the app, and entered it in order to use it. They were given a booklet of trigger images, a checklist of the steps necessary to complete the task, a problem card, and all other materials necessary for completing the mathematics tasks (writing utensils and calculator). The participants received the initial prompt “Please solve the problem” and were expected to do so, with the segments showing the individual steps (VP) available to them. The researcher collected data both on the participants’ performance of the skills and on which videos they watched. Each data point represented an opportunity to complete the selected task. There were an average of 20 steps possible for each problem; recipe adjustment had 18 steps, calculation of unit prices had 24 steps, and calculation of salary had 16 steps. Each intervention session lasted 30 min on average. In order to move on to the maintenance phase, the participants had to perform a task at least 5 times (five sessions of intervention) with at least 80% accuracy (the final result still being correct) during at least three consecutive sessions.

Maintenance

Maintenance data were collected in 1-week increments after the intervention phase was completed. Each skill was tested for maintenance for at least one session (1 week). The maintenance phase procedures were identical to those used in the baseline phase (e.g., no videos or other resources).

Interobserver agreement

Each session of each phase (baseline, intervention, and maintenance) was recorded with a camera, focused mainly on the problem and the solving process. At the end of the study, 35% of the recorded sessions of each phase for each participant (randomly selected) were watched and scored again by a second researcher with experience in special education who was trained in the data collection procedures. Percentages of agreement were then calculated dividing the number of agreements by the total number of agreements plus disagreements. An overall agreement of 98.7% was achieved. Interobserver agreement results for each participant broken down by task and phase can be found in Table 2.

Table 2.

Interobserver Reliability by Participant and Phase.

Participants	Recipe Adjustment (%)			Calculation of Salary (%)			Calculation of Unit Prices (%)
Participants	BL	INT	MT	BL	INT	MT	BL	INT	MT
Sophie	100	99	100	98	100	100	96	100	100
Ella	100	99	100	96	100	100	97	98	100
Jane	100	99	100	87	100	100	97	100	100

BL = baseline; INT = intervention; MT = maintenance.

Results

This intervention used VP and a checklist delivered through an AR platform to teach functional mathematics skills to three adults with ID. The omnibus results will be presented first followed by individual participant results. In baseline, across the three skills and across each participant, the average percent correct was 16%. In the intervention phase, the overall average accuracy across participants and skills was 98% (range of 82–100). Visual analysis of the graphs for all the participants and skills determined that an immediate change in level occurred when the intervention was implemented on all nine graphs. Using the split-middle method (White & Haring, 1980), the trend in the intervention phase was level or increasing for all nine of the intervention phases. The data in the intervention phase were deemed to be stable, with all nine graphs having 80% of the data being within a 25% range of the median value (Barton et al., 2018). The overall Tau-U combined and weighted Tau-U score for the intervention and maintenance phases across skills was 1.00.

Percentage of Steps Completed Correctly

The first research question addressed the relationship between the dependent variable (percentage of steps completed correctly) and the independent variable (the intervention) for three targeted tasks. The individual results for the selected tasks follow.

Sophie

The results for Sophie’s performance are reported in Table 3 and Figure 1. During the baseline phase of the recipe adjustment task, Sophie completed none of the steps correctly over six sessions. During the intervention phase of this task, Sophie’s performance accuracy increased to 28% immediately after the intervention was introduced, reached 100% accuracy after three sessions, and remained at 100% for six consecutive sessions.

Table 3.

Average Individual and Overall Percentages of Performance Accuracy.

Participants	Adjustment of Recipe (%)			Calculation of Salary (%)			Calculation of Unit Prices (%)
Participants	BL	INT	MT	BL	INT	MT	BL	INT	MT
Sophie	0	82	100	30	100	100	17	100	100
Ella	0	100	100	39	96	100	4	100	100
Jane	0	99	100	26	100	100	30	100	100
Overall	0	94	100	32	99	100	17	100	100

BL = baseline; INT = intervention; MT = maintenance.

Figure 1.

Percentage of steps completed correctly by Sophie.

During the baseline phase of the calculation of unit prices task, Sophie completed an average of 17% of the steps correctly, with a range of 13%–25% over nine sessions, with the last six stabilizing at 13%. During the intervention phase of calculating unit prices, Sophie’s performance accuracy increased to 100% immediately after the intervention was introduced, maintaining at 100% for eight consecutive sessions.

For the calculation of salary task, Sophie’s baseline performance was an average of 30% of the steps completed correctly, with a range of 25%–62% over 12 sessions, with the last four stabilizing at 38%. Her performance accuracy during the intervention phrase increased to 100% immediately after the intervention was introduced and maintained at 100% for six consecutive sessions.

Ella

Ella’s results are reported in Table 3 and Figure 2. During baseline for the calculation of unit prices, Ella completed an average of 4% of the steps correctly; her range was 0–13% over six sessions, the last three stabilizing at 6%. During intervention, Ella’s accuracy at calculating unit prices increased to 100% immediately after the intervention was introduced, remaining at 100% for eight of the nine maintenance sessions, with seven consecutive sessions at 100%.

Figure 2.

Percentage of steps completed correctly by Ella.

During baseline of the calculation of salary task, Ella completed an average of 39% of the steps correctly, ranging from 13% to 62% over nine sessions, with the last three stabilizing at 62%. During the intervention phase, Ella’s accuracy at calculating salary increased to 75% immediately after the intervention was introduced, reaching 100% accuracy after two sessions and remaining at 100% for five consecutive sessions.

During baseline for the task of recipe adjustment, Ella completed none of the steps correctly over 12 sessions. During the intervention phase, Ella was able to adjust recipes at 100% accuracy immediately after the intervention was introduced, remaining at 100% for six consecutive sessions.

Jane

Jane’s performance results are reported in Table 3 and Figure 3. During the baseline phase of the calculation of salary task, Jane completed an average of 26% of the steps correctly, with a range of 25–38% over six sessions, with five of those stabilizing at 25% with only one exception at 38%. During the intervention phase of this task, Jane’s performance accuracy increased to 100% immediately after the intervention was introduced and remained at 100% for seven consecutive sessions.

Figure 3.

Percentage of steps completed correctly by Jane.

During the baseline phase of the recipe adjustment task, Jane completed none of the steps correctly over nine sessions. During the intervention phase, Jane was able to adjust recipes at 100% accuracy immediately after the intervention was introduced and remained at 100% for six of the eight sessions and with five consecutive sessions at 100%.

During the baseline phase of the calculation of unit prices, Jane completed an average of 30% of the steps correctly, ranging from 13% to 62% over 12 sessions with four out of the last six stabilizing at 37%. During the intervention phase, Jane’s performance accuracy at calculating unit prices increased to 100% immediately after the intervention was introduced and remained at 100% for six consecutive sessions.

Task Maintenance

The purpose of the second research question was to determine the ability of the participants to maintain the skills without using the intervention. For each target task, maintenance data were collected in 1-week increments after the intervention phase was completed. As reported in Table 2, the overall average percentage of steps completed correctly during the maintenance phase was 100% across tasks. The individual maintenance results for the selected tasks are as follows.

Sophie

The results for Sophie’s maintenance performance are reported in Table 2 and Figure 1. During the maintenance phase of the recipe adjustment task, Sophie completed 100% of the steps correctly over three sessions of the 1-week increments. For the calculation of unit prices task, she completed 100% of the steps correctly over two sessions, and for the calculation of salary task, she completed 100% of the steps correctly during one session.

Ella

Ella’s maintenance results are shown in Table 2 and Figure 2. For the calculation of unit prices task, Ella maintained 100% of the steps correctly over three sessions in the 1-week increments. On the calculation of salary task, she maintained 100% of the steps correctly over two sessions, and on the recipe adjustment task, she maintained 100% of the steps correctly during one session.

Jane

Maintenance results for Jane are reported in Table 2 and Figure 3. In calculating salary, Jane completed 100% of the steps correctly over three sessions in the 1-week increments. In adjusting recipes, she completed 100% of the steps correctly over two sessions, and in calculating unit prices, she completed 100% of the steps correctly during one session.

Posttest

The posttest, similar in difficulty to the pretest, which each participant had completed with 0% accuracy was administered after the last maintenance session, confirmed the performance accuracy percentage shown in the maintenance phase. All three participants completed the posttest with 100% accuracy.

Social Validity

The purpose of the third research question was to determine whether the participants perceived the skills taught and the intervention used as socially valid and useful overall. All the participants were consistent in responding positively to the questionnaire. They all confirmed they liked learning the skills because these skills were new and involved things that were part of their daily lives, like shopping and cooking; their favorite task was the recipe adjustment because it involved food, which they like. All the participants enthusiastically said that they planned to use the new skills, especially while cooking and shopping, and Jane reported she had already started adjusting the number of servings on recipes while cooking with her family.

The participants unanimously reported that they enjoyed using the iPad to learn because it was different from usual instruction and more fun. Sophie said using the iPad made her feel smart and accomplished because she could learn things by herself. Ella noted that watching the videos was easier than listening to a person explaining a task because the instruction was always the same, which helped her remember it better. All of the participants said that the videos used clear steps, which helped their learning, that the iPad and the HP-Reveal app were easy to use, and that already knowing how to use an iPad helped. Ella and Jane, however, mentioned they had difficulty with double tapping the videos to make them full screen once they had been triggered.

All the participants became very excited in telling the researcher that they would like to learn other things using the videos/pictures on the iPad. Ella mentioned she would like to learn about science, specifically about the ocean and the creatures in it, through the same intervention packet. Jane said she loved theater and would like to learn through videos how to make props. Sophie explained she would like to learn more about cooking in general and about baking techniques specifically.

All of them felt that the intervention was good but remarked that math was not their favorite subject and that the problems were difficult in the beginning, which they did not like. The three participants provided some suggestions to improve the intervention and the materials: Mainly they recommended that the problems include pictures to help them understand the text better, and they said they would appreciate having the problems written in bigger font.

Tau-U Numbers and p Values Results

The average Tau-U effect size between baseline and intervention was 1.0 for all three tasks. All values denoted a clear difference in the students’ performance between the baseline phase and the intervention phase, supporting the hypothesis of intervention functionality.

The combined p value was .0010 for the recipe adjustment task, .0009 for the calculation of unit prices task, and .0016 for the calculation of salary task. All values denoted a high relevance in regard to the data collected and the study findings because they showed a significant difference between students’ performance between baseline and intervention phases.

Discussion

The purpose of this study was to determine the effectiveness of using VP delivered via AR to teach transition-related math skills to participants with ID. The results of the study extend the abundant literature supporting the effectiveness of VP specifically and VM in general for teaching academically relevant materials and skills to individuals with disabilities (Burton et al., 2013; Kellems et al., 2016). This study also adds to prior evidence that iPads can be used as teaching tools to help adults with disabilities learn academic skills (Kagohara et al., 2013; Kellems & Morningstar, 2012). The findings of this study also support the claim that use of AR has great potential in the field of education (Wu et al., 2013), including its effectiveness in providing a platform for delivering formal content such as mathematics (Sommerauer & Muller, 2014). Further research is needed to determine its effectiveness with different subjects and groups of people with various disabilities and needs.

Maintenance results of 100% performance across tasks and participants stand as evidence of the durability of the intervention’s effects on maintaining the learned skills. Such results are consistent with the findings of Kellems et al. (2017) and Kellems et al. (2016) in demonstrating that individuals with disabilities can learn and maintain skills taught to them via VM over extended periods of time. Such results also suggest that the intervention packet was effective both in guiding the participants through completion of the targeted tasks through the step-by-step videos and also in teaching them the steps that they were later able to perform on their own without the aid of the intervention materials. The fact that the percentage of performance accuracy did not decrease with the withdrawal of the intervention supports the hypothesis of its functionality in teaching academic material such as math and also in helping the participants retain the information.

The social validity findings of the study are similar to those of Kellems et al. (2016) in that both studies reported that the participants wanted to learn about other subjects using the same intervention which they perceived as easier to use, more entertaining, and more engaging than in vivo instruction. The students’ enthusiastic responses to the questionnaire indicated that the intervention was successful not only statistically but in its perception by participants as a functional and effective way of learning. These findings are significant for researchers and practitioners as they provide support that VP delivered via AR is a socially valid intervention that individuals with disabilities enjoy using.

Recent literature has indicated the need for conducting additional research on effective ways of teaching math-related skills to individuals with disabilities (Browder et al., 2008). The present study adds to the literature by providing information about an emerging intervention used to teach math content to individuals with disabilities.

Limitations

A number of limitations in this study should be addressed in future research. The technology used (HP-Reveal), which was developed in 2011, is still being refined. Also, only three participants were selected for the study, all with the ID classification of similar ages, which did not allow for comparison with other disabilities. Another weakness involved materials and equipment: Creating the materials was time-consuming, and several pieces of equipment had to be set up at the beginning of each session. More efficient use of resources could be sought. As a result of the multiple baseline design, students learned some tactics from the first and second tasks which they started using in the second and third tasks, causing the baseline accuracy percentages to increase moderately, though they remained low enough for the intervention to be necessary and meaningful.

Another limitation to consider is that the intervention package consisted of a video component and a checklist. The current study did not collect any data as to which part of the intervention package was the most salient for the participant. Potentially the intervention could have been as effective with only a single component. Data related to the procedural reliability of the implementation for the intervention were not collected which raises the question as to the fidelity of implementation of the intervention. One additional limitation to consider is that the intervention was delivered in a distraction-free environment. The positive results in some part may have had to do with the environment.

Suggestions for Future Research

Future research should focus on delivering the intervention packet to participants of different ages, with different disabilities, with different levels of mathematical ability, and in different settings. Future research should also explore how the use of this or a similar intervention including AR and VP might be used to teach a wider range of math skills (geometry, etc.) and other subjects such as English, science, and art. Furthermore, additional research could focus on determining which electronic devices are the most functional and relevant in teaching target skills.

Implications for Practitioners

To become contributing members of society, individuals with disabilities should be proficient in math skills that will allow them to access competitive employment and to live as independently as possible. Like their nondisabled peers, individuals with disabilities should be systematically taught such skills in ways that will help them learn effectively and retain the knowledge and skills for extended periods of time. Practitioners who work with individuals with disabilities should rely on evidence-based practices to teach these individuals whenever possible.

While creation of the AR-based intervention may take some additional up-front time results from this study and the emerging body of research suggest that it may be worth the extra time invested based on the significant gains made by students. It may also be beneficial to use an AR-based strategy for students who may not respond to traditional VM interventions. However, in some cases, the extra time warranted for creating an AR-based intervention may not be worth it if a traditional VM intervention would yield the same results.

The intervention used in the present study is anchored in the extensive research available on the effectiveness of VM and VP; thus, these strategies can be considered valid tools to teach individuals with disabilities relevant and useful math skills. Less research is available on the use of AR. While this study used AR as a platform for delivering the video prompts, published research that utilizes AR in this way and in other ways all indicate its potential, especially in the field of education (Wu et al., 2013).

Conclusion

This study sought to examine use of an AR intervention package in teaching math skills pertinent to the transition and vocational development of adult individuals with ID. A functional relationship between the dependent and the independent variables was established by comparing students’ performance during baseline and after the introduction of the intervention package. All students showed a fast and significant improvement in their performance after receiving the intervention and also maintained the learned skills after removal of the intervention. Further research is needed to provide more evidence of the effectiveness of AR in teaching individuals with disabilities not only math-related content but also other subject skills.

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

Ryan O. Kellems

References

Allen

K. D.

Wallace

D. P.

Renes

Bowen

S. L.

Burke

R. V.

(2010). Use of video modeling to teach vocational skills to adolescents and young adults with autism spectrum disorders. Education & Treatment of Children, 33, 339–349. https://doi.org/10.1353/etc.0.0101

Banda

D. R.

Dogoe

M. S.

Matuszny

R. M.

(2011). Review of video prompting studies with persons with developmental disabilities. Education & Training in Autism & Developmental Disabilities, 46, 514–527.

Barton

E. E.

Lloyd

B. P.

Spriggs

A. D.

Gast

D. L.

(2018). Multiple baseline and multiple probe designs. In Ledford

Gast

(Eds.), Single case research methodology: Applications in special education and behavioral sciences. (3rd ed., pp. 239–281). Routledge.

Bellini

Akullian

(2007). A meta-analysis of video modeling and video self-modeling interventions for children and adolescents with autism spectrum disorders. Exceptional Children, 73, 264–287. https://doi.org/10.1177/001440290707300301

Browder

D. M.

Spooner

Ahlgrim-Delzell

Harris

A. A.

Wakemanxya

(2008). A meta-analysis on teaching mathematics to students with significant cognitive disabilities. Exceptional Children, 74(4), 407–432. https://doi.org/10.1177/001440290807400401

Burton

C. E.

Anderson

D. H.

Prater

M. A.

Dyches

T. T.

(2013). Video self-modeling on an iPad to teach functional math skills to adolescents with autism and intellectual disability. Focus on Autism and Other Developmental Disabilities, 28(2), 67–77. https://doi.org/10.1177/1088357613478829

Cannella-Malone

Fleming

Chung

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, 144–153. https://doi.org/10.1177/1098300710366593

Cannella-Malone

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, 344–356.

Charlop-Christy

M. H.

Freeman

K. A.

(2000). A comparison of video modeling with in vivo modeling for teaching children with autism. Journal of Autism and Developmental Disorders, 30(6), 537–552.

10.

Cihak

D. F.

Moore

E. J.

Wright

R. E.

McMahon

D. D.

Gibbons

M. M.

Smith

(2016). Evaluating augmented reality to complete a chain task for elementary students with autism. Journal of Special Education Technology, 31, 99–108. https://doi.org/10.1177/0162643416651724

11.

Corbett

B. A.

Abdullah

(2005). Video modeling: Why does it work for children with autism? Journal of Early and Intensive Behavior Intervention, 2(1), 2.

12.

Gardner

Wolfe

(2013). Use of video modeling and video prompting interventions for teaching daily living skills to individuals with autism spectrum disorders: A review. Research and Practice for Persons with Severe Disabilities, 38, 73–87. https://doi.org/10.1177/0162643416651724

13.

Gast

D. L.

Loyd

B. P.

Ledford

J. R.

(2018). Synthesis and meta-analysis of single case research. In Ledford

Gast

(Eds.), Single case research methodology: Applications in special education and behavioral sciences. (3rd ed., pp. 365–391). Routledge.

14.

Kagohara

D. M.

van der Meer

Ramdoss

O’Reilly

M. F.

Lancioni

G. E.

Davis

T. N.

Rispoli

Lang

Marschik

P. B.

Sutherland

Green

V. A.

Sigafoos

(2013). Using iPods® and iPads® in teaching programs for individuals with developmental disabilities: A systematic review. Research in Developmental Disabilities, 34(1), 147–156. https://doi.org/10.1016/j.ridd.2012.07.027

15.

Kellems

R. O.

Cacciatore

Osborne

(2019). Using an augmented reality–based teaching strategy to teach mathematics to secondary students with disabilities. Career Development and Transition for Exceptional Individuals, 42(4), 253–258. https://doi.org/10.1177/2165143418822800

16.

Kellems

R. O.

Eichelberger

Cacciatore

Jensen

Frazier

Simons

Zaru

(2020). Using video-based instruction via augmented reality to teach mathematics to middle school students with learning disabilities. Journal of Learning Disabilities. https://doi.org/10.1177/0022219420906452

17.

Kellems

R. O.

Frandsen

Hansen

Gabrielsen

Clarke

Simons

Clements

(2016). Teaching multi-step math skills to adults with disabilities via video prompting. Research in Developmental Disabilities, 58, 31–44. https://doi.org/10.1016/j.ridd.2016.08.013

18.

Kellems

R. O.

Morningstar

M. E.

(2012). Using video modeling delivered through iPods to teach vocational tasks to young adults with autism spectrum disorders. Career Development and Transition for Exceptional Individuals, 35, 155–167. https://doi.org/10.1177/0885728812443082

19.

Kellems

R. O.

Rickard

T. H.

Okray

D. A.

Sauer-Sagiv

Washburn

(2017). iPad® video prompting to teach young adults with disabilities independent living skills: A Maintenance study. Career Development and Transition for Exceptional Individuals, 41(3), 175–184. https://doi.org/10.1177/2165143417719078

20.

Kena

Hussar

McFarland

de Brey

Musu-Gillette

Wang

Zhang

Rathbun

Wilkinson-Flicker

Diliberti

Barmer

Bullock Mann

Dunlop Velez

(2016). The condition of education 2016. NCES 2016-144. National Center for Education Statistics. https://eric.ed.gov/?id=ED565888

21.

Kennedy

C. H.

(2005). Single-case designs for educational research. Pearson Education.

22.

King

S. A.

Lemons

C. J.

Davidson

K. A.

(2016). Math interventions for students with autism spectrum disorder. Exceptional Children, 82, 443–462. https://doi.org/10.1177/0014402915625066

23.

Kratochwill

T. R.

Hitchcock

Horner

R. H.

Levin

J. R.

Odom

S. L.

Rindskopf

D. M

Shadish

W. R.

(2010). Single case designs technical documentation. In What works clearinghouse: Procedures and standards handbook (Version 2.0). What Works Clearinghouse. http://ies.ed.gov/ncee/wwc/pdf/wwc_procedures_v2_standards_handbook.pdf

24.

Lin

Chai

Wang

Chen

Liu

Chen

Huang

(2016). Augmented reality in educational activities for children with disabilities. Displays, 42, 51–54. https://doi.org/10.1016/j.displa.2015.02.004

25.

McMahon

D. D.

Cihak

D. F.

Wright

R. E.

Bell

S. M.

(2016). AR for teaching science vocabulary to postsecondary education students with ID and autism. Journal of Research on Technology in Education, 48, 38–56. https://doi.org/10.1080/15391523.2015.1103149

26.

McMahon

D. D.

Smith

C. C.

Cihak

D. F.

Wright

Gibbons

M. M.

(2015). Effects of digital navigation aids on adults with intellectual disabilities: Comparison of paper map, Google Maps, and augmented reality. Journal of Special Education Technology, 30, 157–165. https://doi.org/10.1177/0162643415618927

27.

No Child Left Behind Act of 2001, P.L. 107-110, 20 U.S.C. § 6319 (2002).

28.

Patton

J. R.

Cronin

M. E.

Bassett

D. S.

Koppel

A. E.

(1997). A life skills approach to mathematics instruction: Preparing students with learning disabilities for the real-life math demands of adulthood. Journal of Learning Disabilities, 30, 178–187. https://doi.org/10.1177/002221949703000205

29.

Schalock

R. L.

Borthwick-Duffy

S. A.

Bradley

V. J.

Buntinx

W. H.

Coulter

D. L.

Craig

E. M.

Gomez

S. C.

Lachapelle

Luckasson

Reeve

Shogren

K. A.

Snell

M. E.

Spreat

Tasse

M. J.

Thompson

J. R.

Verdugo-Alonso

M. A.

Wehmeyer

M. L.

Yeager

M. H.

(2010). Intellectual disability: Definition, classification, and systems of supports. American Association on Intellectual and Developmental Disabilities.

30.

Schulte

A. C.

Stevens

J. J.

(2015). Once, sometimes, or always in special education: Mathematics growth and achievement gaps. Exceptional Children, 81, 370–387. https://doi.org/10.1177/0014402914563695

31.

Sommerauer

Muller

(2014). Augmented reality in informal learning environments: A field experiment in a mathematics exhibition. Computers & Education, 79, 59–68. https://doi.org/10.1016/j.compedu.2014.07.013

32.

Taber-Doughty

Bouck

E. C.

Tom

Jasper

A. D.

Flanagan

S. M.

Bassette

(2011). Video modeling and prompting: A comparison of two strategies for teaching cooking skills to students with mild intellectual disabilities. Education and Training in Autism and Developmental Disabilities, 46, 499–513.

33.

U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics (2015). https://nces.ed.gov/

34.

Walton

K. M.

Ingersoll

B. R.

(2013). Improving social skills in adolescents and adults with autism and severe to profound intellectual disabilities: A review of the literature. Journal of Autism and Developmental Disorders, 43, 594–615. https://doi.org/10.1007/s10803-012-1601-1

35.

Wei

Lenz

K. B.

Blackorby

(2013). Math growth trajectories of students with disabilities: Disability category, gender, racial, and socioeconomic status differences from ages 7 to 17. Remedial & Special Education, 34, 154. https://doi.org/10.1177/0741932512448253

36.

White

O. R.

Haring

N. G.

(1980). Exceptional teaching: A multimedia training package. Merrill.

37.

H.-K.

Lee

S. W.-Y.

Chang

H.-Y.

Liang

J.-C.

(2013). Current status, opportunities and challenges of augmented reality in education. Computers & Education, 62, 41–49. https://doi.org/doi:10.1016/j.compedu.2012.10.024

38.

Yakubova

Hughes

Hornberger

(2015). Video-based intervention in teaching fraction problem-solving to students with autism spectrum disorder. Journal of Autism and Developmental Disorders, 45, 2865–2875. https://doi.org/10.1007/s10803-015-2449-y