Effects of video modeling on addition word-problem performance of students with autism spectrum disorder

Abstract

Many students with autism spectrum disorder (ASD) experience academic challenges and difficulties. These struggles are especially pronounced in mathematics with students with ASD performing significantly below than their peers without disabilities on measures of mathematical performance. The current study used a single case experimental design with concurrent multiple probe across students to investigate the effects of a point-of-view video modeling (POVM) intervention on accuracy of addition with regrouping word problems. The participants were three secondary grade level students with ASD. Findings showed that each student demonstrated considerable improvement during intervention over baseline levels, and subsequently sustaining their performance through the maintenance phase. Limitations, implications for practitioners and future research directions are presented.

As a spectrum disorder, autism encompasses a range of developmental disabilities and characteristics (King et al. 2016). While the characteristics most commonly associated with autism spectrum disorders (ASD) are the social, communicative and behavioral traits identified in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) definition of ASD, students with ASD also experience variations in academic challenges. These variations may result in what Keen et al. (2016) referred to as uneven patterns of achievement, which may be masked when looking at individuals with ASD collectively rather than at the student level. Wei et al. (2015) reported that students with ASD demonstrate heterogeneous academic profiles, with approximately one third of students displaying overall poor academic performance. The variability of academic performance extends to mathematics (King et al. 2016).

In an analysis of data from the Special Education Elementary Longitudinal Study (SEELS), Wei et al. (2013) reported that the average mathematics scores for students with ASD was one standard deviation below the average for peers without disabilities. For a subset (20%) of students in their study who were regarded as having an academic profile consistent with hypercalculia (higher than expected achievement in mathematics). Their performance on calculations was similar to the national average, but performance on applied problems was considerably lower. Researchers have raised concerns that the mathematics growth trajectories of students with ASD project less substantial gains than peers without ASD (Keen et al. 2016, Wei et al. 2013). Students with ASD struggle academic issues like mathematical word problems that are language-based and require skills in executive function. Given the poor mathematical performance of many students with ASD and the concerned related to growth trajectories, it is important to examine the collection of intervention research targeting the mathematical performance of students with ASD.

Previous research on mathematical interventions for students with ASD

Recent research has combined different components such as modeling, manipulatives, and task analysis to improve addition and subtraction problem solving performance of students with ASD. In the study conducted by Root et al. (2017), researchers employed modified schema-based (MSBI) instruction with manipulatives to teach single- and double-digit addition and subtraction computational skills to students with ASD and found a functional relation between MSBI and problem solving. In another study, participants were taught to solve addition/subtraction word problems using explicit instruction, a task analysis, a graphic organizer, and a visual support (Root et al. 2018). Results indicated that all participants were able to discriminate problem types and solve problems. In the study conducted by Root and Browder (2019), the effects of MSBI is evaluated on the problem solving skills of students with ASD/ID. Results indicated that there is a functional relation between the intervention and mathematical word problem solving.

In addition, there are four comprehensive reviews of the literature that focused on or included mathematics interventions for students with ASD have been published. Collective findings from these publications provide guidance for researchers and educators responsible for providing mathematics-targeted interventions to students ASD. First, Spencer and her colleagues (2014) conducted a broad synthesis of academic interventions that included five studies targeting mathematics performance of elementary and mixed-grade level students, and very few secondary grade level students. The authors reported that all five interventions including visual supports, technology-based instruction, concrete representation, direct instruction and behavioral interventions were effective. Second, extending the findings from Spencer et al. (2014), Hart-Barnett and Cleary (2015) reported effectiveness for the intervention designs included the use of visual representations (e.g. video self-modeling, concrete representation) and cognitive strategies (e.g. counting on strategy, response repetition) for elementary and secondary students with ASD. Third, a meta-analysis conducted by King et al. (2016) reported 21 studies had a moderate effect size for mathematical interventions and aligned with the findings of both Spencer et al., and Hart-Barnett and Cleary. Finally, the results of a systematic analysis by Gevarter et al. (2016) also found the effectiveness of several mathematics interventions for students with ASD. The collective findings indicated that one of the most notable interventions across these reviews is video-based instruction (VBI).

Video-based instruction

VBIs have a historical foundation of empirical support for social-communicative and behavioral skills for students with ASD (Bennett et al. 2017) and a rapidly emerging empirical base for use to teach mathematics. VBI research focusing on mathematics became more visible in the past decade with Hart and Whalon (2008) and Burton et al. (2013) near the beginning of this more recent wave of research that utilized video modeling to target mathematics skills (i.e. dependent measure evaluated counting and giving change) for students with ASD and offered a new and promising intervention for academic interventions. Research documents VBI as an effective medium to teach mathematics to students across formal schooling years, from early elementary students (e.g. Jowett et al. 2012, Knight et al. 2018, Yakubova et al. 2015) to young adults (Kellems et al. 2016). Clearly, a body of research demonstrating the positive impact of VBIs for students with ASD is growing and an important option for consideration by educators.

The term VBI encompasses several varieties of the video intervention. Video-modeling shows a video where a peer or adult serves as the model and demonstrates the skill or behavior. Video-self modeling is similar to video modeling, but with the unique variation that the individual receiving the intervention is recorded modeling the skill. Video-prompting displays a multi-stepped task divided into steps and is watched in portions as the individual completes each step. Point-of-view video modeling (POVM) shows completion of a skill or task from the perspective of the student, or individual completing the task. Variations of the VBI occur, also, such as VP-POVM or POVM-SM.

Including a VBI has several advantages. First, the video draws viewers’ attention and focus to the targeted task, limiting external stimuli that may distract from the lesson objective (Hughes and Yakubova 2016, Mason et al. 2013). Second, unlike live instruction, students are able to watch and rewatch the video as often as needed (Yakubova et al. 2015), allowing for consistent repetition of language and instruction. Third, videos can depict instructional components that have established empirical support to teach concepts and procedures in mathematics (Root et al. 2018), such as multiple representations (e.g. Bouck et al. 2014, Weng and Bouck 2014, Yakubova et al. 2015). Similarly, VBI may be coupled with behavioral components that have established histories of success and can be scaffolded to provide more or less support as needed, such as system of least prompts (Knight et al. 2018) or gradual fading (Jowett et al. 2012).

Across the different types of VBI, POVM may organically lend itself to academic instruction for mathematics. POVM visually records the task from an angle that replicates the angle from which the student would complete. As such, it shows video of the model’s hands completing the academic task accompanied by an audio annotation, or instruction, of the task. The video view allows for optimal visual focus on the specific actions involved with the task. POVM has been used to teach fundamental skills, such as counting (Jowett et al. 2012), addition, subtraction and number comparison (Yakubova et al. 2016), as well as more complex skills, such as solving word problems that require subtracting fractions with unlike denominators (Yakubova et al. 2016). Given the mathematical struggles of students with ASD and the emerging yet promising use of VBI, it is important for researchers to continue to examine and refine the use of VBI to improve the mathematical performance of students with ASD.

Purpose

The purpose of this research is to add to and extend the current evidence-base for VBI, and specifically POVM, by evaluating the effects of POVM intervention on the word problem solving performance of students with ASD. Additionally, this research extends the evidence base of VBI to students with ASD in Turkey. The primary research questions are:

1. What are the effects of a POVM intervention using a five-step checklist to solve word problems involving addition with regrouping for students with ASD?

2. What are the effects of a POVM intervention on maintenance of solving word problems involving addition with regrouping without the reminder of five-step checklist or the intervention?

Methods

Participants

Three male students, Alex, Tom, and Sam, attending the same private special education school in Turkey participated in this study. Ethical approval was granted by the university ethical board. Parents of children provided written consents for their children to participate in the study. All procedures performed in studies involving human participants were in accordance with the ethical standards of the Helsinki declaration. Students were selected based on the following criteria: (a) received the diagnoses from child psychiatrists who work for state hospitals and/or university hospitals meeting the ASD diagnostic criteria according to Diagnostic and Statistical Manual of Mental Disorders 5^th edition (DSM-V; APA 2013), (b) deficiency in problem solving with two-digit plus one-digit addition problems with regrouping, (c) demonstrated reading comprehension of basic word problems, (d) no prior experience with video modeling instruction, (e) no vision or hearing impairments and (f) agreeing to participate in the study and parental permission. Four students started the study intervention, but one student was dropped out of the study due to continued absences.

Participants received their diagnosis from a pediatric psychiatrist at a private hospital. All students were diagnosed with ASD and scored in the average range of autism on Gilliam Autism Rating Scale-2 (GARS-2; Gilliam 2006). GARS-2 is used for the purpose of identifying individuals with ASD. Students with disabilities in the private special education school receive face-to-face and/or group education as well as rehabilitation or therapy services. Instruction in these services include training in academic, social, communicational, and daily living skills according to each students' individualized education program (IEP). All students participated in individual instruction for mathematics.

Alex, 15 years old, has been attending a fully integrated private school since third grade. His IQ score is 81 (WISC-R; Wechsler 1974) and demonstrated average performance in reading with no significant academic concerns. His teacher described him as a good reader and parents noted he experienced significant difficulties in interaction especially in group situations. Tom, 15 years old, has been attending a private special education school since five years old. His IQ score is 72 (WISC-R) and reading level falls within the 5^th–6^th grade. His parents and teachers described him as an energetic boy with average academic skills who enjoys reading. His teacher noted his stimming behaviors such as hand-flapping result in off-task behavior during academic situations. Sam, 16 years old, has been attending the same private special education school since second grade. His IQ score is 70 (WISC-R) and below average in all academic areas, but especially mathematics. He is able to read in a slow manner and calculate addition problems without regrouping. His mother described him as showing minimal interest in academic areas and runs away often when asked to complete assignments.

Mathematics learning objectives

All students were receiving instruction following the state-supervised national curriculum. At the time of the study (beginning of school year), students had received instruction involving basic single digit addition facts or two digits plus one-digit addition without regrouping in isolation followed by those skills in word problems following the current mathematics scope sequence. The weekly hours of mathematics support that they receive is approximately 2 h. Students had not yet received any instruction on extending the addition computation to include regrouping or word problems requiring regrouping. According to the students’ teacher, all three participants had demonstrated proficiency with one-digit plus one-digit and two-digit plus one-digit addition problems but had deficiencies in problem solving with two-digit plus one-digit addition problems with regrouping. We verified their current performance using a screener comprised of addition problems without regrouping and with regrouping. The screener required students to complete 10 one-digit plus one-digit and two-digit plus one-digit addition problems. Results indicated Alex, Tom and Sam were able to solve the addition problems without regrouping with 100%, 100%, 80% criteria, respectively. Since it is the next step in the curriculum, we identified addition with regrouping word problems as the target skill for this study.

Using the classroom mathematics curriculum, we created a bank of 30-word problems involving addition with regrouping that matched the type of word problems in the curriculum. The classification of all the word problems were same and they were ‘get more’ change problems where result is unknown in the final position (Okamoto 1996). The classroom teacher reviewed all problems and confirmed they aligned properly with the math curriculum. All word problems involved addition with regrouping solved either with one step or two steps. The 1-step problems only contained two numbers and the 2-step word problems contained three numbers. We first taught the students the 1-step and then moved to the 2 step word problems. Table 1 contains an example of both word 1 and 2-step word problems in Turkish and English. The word problems were presented in Turkish.

Table 1.

1-and 2-step word problems involving addition with regrouping.

	Turkish	English
1-step	Ashley’nin 15 tane kurabiyesi var. Kardeşi Andy ona 8 tane daha kurabiye verdi. Ashley’nin şimdi kaç tane kurabiyesi var?	Ashley has 15 cookies and her brother Andy gave her 8 more cookies. How many cookies does Ashley have now?
2-step	Zeynep’in 28 tane kalemi var. Babası, Zeynep’in kalemlerine 6 kalem daha ekledi. Daha sonra Zeynep arkadaşından 7 tane daha aldı. Bu durumda Zeynep’in kaç kalemi olur?	Zeynep has 28 pencils. Her dad gives 6 more pencils to her. Zeynep gets 7 more from her friend. At the end, how many pencils does Zeynep have?

Settings

The study conducted in a private school for students with disabilities. The school is located in a large urban city in the Northern part of Turkey and serves children in pre-kindergarten through age 23. The school serves approximately 300 students and the average students attend 2–3 days per week for approximately 2–4 h. Instruction is typically delivered in 1:1 environments.

Teachers provided mathematics instruction following a curriculum and students Individualized Education Program (IEP) goals. Both baseline and intervention sessions occurred during the allotted time for the mathematics period in the special education school. All video-modeling intervention sessions were delivered in a separate classroom to limit distractions of other students in their classroom. Maintenance sessions occurred in each student’s regular mathematics period. In all sessions, there was a table and two chairs, one for the teacher and one for the student. The teacher and student sat at their chairs facing each other while a student was solving problems or watching the video. Both baseline and intervention sessions were conducted in a 1:1 environment while maintenance sessions were conducted in their classrooms.

Materials

Video-modeling clip

The video clip was created by the mathematics classroom teacher and lead investigator and recorded from a first-person point-of-view. The teacher served as the adult hands writing out the addition problems and verbally narrating the five-step checklist in the students’ native language, Turkish. The verbal narration included the transitioning between the problem solving checklist such as ‘I completed this step and I am moving to the next step’. The video clip employed explicit instruction and depicted exactly what the child’s viewpoint should look like: two hands, pencil and a paper. The video clip including the solution of a problem lasted 3 min and 9 s.

Five-step problem solving checklist

A five-step checklist was developed by the researchers to scaffold students’ learning to solve word problems. The steps included (1) read the problem; (2) represent problem with numbers; (3) determine if there will be regrouping; (4) complete regrouping and (5) compute the answer. Instructional scaffolding is identified as one of 22 High Leverage Practices used to support learning of students with disabilities (McLeskey et al. 2017) and especially appropriate for developing specially designed instruction for mathematics content (Archer and Hughes 2011, Riccomini et al. 2017). The use of checklists to support students with ASD in both behavioral focused tasks as well as academic learning tasks is well documented and recently identified in previous research for students with ASD (Gevarter et al. 2016, Hart-Barnett and Cleary 2015, King et al. 2016, Spencer et al. 2014). The checklist was embedded on the student learning sheet used while viewing the video clip during the intervention. The video clip was uploaded to a tablet PC and students independently accessed the video intervention. Since the intervention was delivered in a separate area, headphones were not used. Students listened through the tablet pc speakers.

Dependent variable

All students used the same materials throughout the study. The lead investigator, in conjunction with the teacher, developed 30-word problems (change) involving two-digit by one-digit addition problems involving regrouping which aligned with the students learning objective and current level of performance. The word problems all involved addition with regrouping solved in 1 or 2 steps. The size of the numbers for the answers were up to 99. The regrouping occurred only in the units digit (e.g. 25 + 7 = 32). Each assessment probe consisted of five word problems. The teacher reviewed all word problems and determined which problems were most appropriate based on the context of the problems.

Reliability

Data for inter-observer agreement and procedural reliability were collected by the lead investigator and a trained graduate student. Data were recorded simultaneously across 33% of all phases of the study. Observers independently recorded the number of addition problems with regrouping solved correctly. Inter-observer agreement was calculated by dividing the number of agreements of student responses by the number of agreements plus disagreements and multiplying by 100%. The mean inter-observer reliability agreement for each student per phase was 100%. Procedural reliability was also recorded simultaneously across 33% of all phases to ensure students had access to a personal electronic device, watched the intervention completely, and used corresponding work sheets. Data confirmed that all materials were set up and intervention procedures accurately implemented 100% of the time.

Experimental design

This study used an experimental single-case research design. We employed a concurrent multiple probe across students (Gast and Ledford 2014). This experimental design sequentially introduces the independent variable across several individuals who exhibit behaviors or skills that occur under similar conditions (Gast and Ledford 2010) and was used to determine the effectiveness of the independent variable on the acquisition and maintenance of word problem involving with addition with regrouping. To meet sufficient scientific rigor requirements research design and generalization of treatment effects for single-case research, we included more than two participants, collected more than three data points in baseline and treatment conditions, and collected maintenance data (NAC 2015).

The POVM intervention was introduced to each participant after five baseline measures. Intervention measures were collected until each participant reached to a preset criteria (80% correct) which represents a mastery or accuracy criterion for the six intervention sessions. After the intervention phase completion, two maintenance data points were collected with a two week delay to determine the extent to which students maintained the acquired skills. Students did not use the intervention before or during maintenance.

Single case researchers have historically relied on visual analysis of the data to identify the evidence of a treatment and strength of the relationship (Gast 2010, Kratochwill et al. 2014). The current research used a four-step visual analysis as the primary method for the data analysis to determine functional relation. First, the baseline data across all three participants was evaluated for pattern and stability. Data were then evaluated for within-phase level, trend, and variability (i.e. consistency of data path). Second, between phase data were examined to identify overlap, immediacy of effect, and consistency of data. A vertical analysis was conducted to substantiate that increases in the dependent variable were due to manipulation of the independent variable. Third, information from all steps were considered to determine existence of a functional relation and strength of evidence. Further, the improvement rate difference (IRD; Parker et al. 2009) was calculated and included as an effect-size metric. The IRD metric was calculated by determining the exceeding data points in treatment phase over all baseline data points divided by the total improved data points in that phase while eliminating the overlapping data points (Parker et al. 2009).

Procedures

The POVM intervention was the independent variable and it was delivered through a table-top computer. The intervention included a step-by-step strategy checklist and verbal narration of the process of solving the addition problems requiring addition with regrouping. Narration included verbally stating the steps illustrated on the left side of the students learning sheets while demonstrating the problem solution. Students were asked to write the sums of the ones next to the computation, divide ones and tens with a line, write ones at ones place under the line and then write tens over the tens place. The five-step checklist guided students through the process of regrouping specifically focusing the students on the new rule relationship for regrouping. The checklist directed students to determine if regrouping was necessary and then to complete regrouping. See Figure 1 for an example of the student learning sheets.

Figure 1.

A student’s learning sheet that includes the written prompts to solve the word problem and a space for the student to solve the problem. In this figure, we translated the Turkish material into English. See Figure 2 for the Turkish version.

The video displayed hands of an adult model writing on the learning sheet using a pencil while verbally narrating the checklist. Students viewing the video intervention were hearing the process and seeing the process in written form juxtaposed to the abstract problem solution (i.e. numbers and symbols 15 + 8). See Figure 2 for a screenshot of the video modeling. Each participant watched the same video independently by using the tablet pc and viewed the POVM instruction at each intervention session.

Figure 2.

Screenshot of the POVM from the intervention. The figure shows the Turkish version that students used. The English translation of the word problem is: Zeynep has 28 pencils. Her dad gives 6 more pencils to her. Zeynep gets 7 more from her friend. At the end, how many pencils does Zeynep have? See Figure 1 for the English translation of the checklist.

Baseline

Baseline included five consecutive sessions per student until a stable state of responses was achieved. No intervention or any assistance was provided during or before baseline sessions and each student worked on the same set of five word problems. Students received the same instruction at the beginning of each baseline session ‘Complete the worksheet that includes five word problems’.

Intervention

Intervention consisted of a minimum of six sessions per student and until at least 80% accuracy of responses was achieved for two consecutive sessions. In all intervention sessions, students watched the POVM intervention before practicing their set of problems for that session. The POVM intervention was ready at the beginning of every session and students clicked ‘start’ to begin the intervention. It was obvious from the baseline sessions the students were familiar with the process of clicking to start the video, following the screen throughout the video etc. and required no additional training. Each student was given the option to view the video again and were permitted if they requested. All the students watched the POVM intervention two times in the first three intervention sessions. Then, no students re-watched the POVM intervention in the remaining sessions. After viewing the POVM intervention, the lead investigator provided the work sheet that included five addition word problems with regrouping. Students received the same instruction used in baseline sessions. The difficulty of the operations was similar throughout the intervention session. After each participant’s completion, the lead investigator marked the questions as correct or incorrect and recorded the percentage correct.

Maintenance

Two weeks after the last intervention session, each student solved a different set of five addition problems with regrouping for two sessions. Students were not provided the checklist or POVM intervention during maintenance sessions. The purpose of collecting maintenance data was to examine continuity of students’ performance of solving word problems accurately.

Social validity

Social validity is an essential concept that deals with the social significance and social appropriateness of the effects of interventions (Gresham 1983, Wolf 1978). For this reason, examining social validity to fulfill concerns of students, parents, teachers and other stakeholders is important as a part of an applied research. In this study, social validity interviews were conducted with each student and their classroom teachers before and after the data collection procedures to identify their opinions and preferences towards using technology/POVM to learn mathematics word problem solving. Informal social validity questions were related to acceptability and usefulness of the intervention and required either yes or no responses with choice of providing their opinions and reasons to each question. Classroom teachers were also asked whether they would be willing to use any type of video-based instruction after witnessing the implementation and its results.

Results

Effects of POVM intervention on the word problem solving performance of students with ASD were evaluated using multiple-baseline design across three participants in baseline, intervention and maintenance phases. Visual analysis showed that each student demonstrated considerable improvement during intervention over baseline levels, subsequently sustaining their performance through the maintenance phase. POVM intervention increased students’ accuracy in performing word problem solving on addition problems with regrouping. Results are presented in Figure 3. IRD scores revealed strong effectiveness of the intervention on students’ word problem solving performance on addition problems with regrouping. Effect size calculations indicated a large effect size of the intervention with an IRD score of 1.0 and nonoverlap of data points between two phases for all participants. IRD score was also obtained as a large effect (1.0) between baseline and maintenance phases on skill maintenance for all participants.

Figure 3.

Percentage of problems answered correctly.

Participant 1

During baseline, Alex completed none of the word problems, resulting in 0% correct responses. At the beginning of the intervention the change in performance was immediate and pronounced. The mean percentage of accurate responding during intervention was 96.67%. The mean level change between baseline and intervention sessions was 96.67%. Alex’s performance remained stable during maintenance where he obtained mean independent performance of 100% on word problem solving.

Participant 2

When solving word problems involving addition with regrouping, Tom’s average baseline performance was 0%. After intervention was initiated, the change in level was immediate and pronounced. The mean percentage of accurate responding during intervention phase was 93.34%. Intervention data were slightly variable with a downward and upward trend. The mean level change between baseline and intervention sessions was 93.34%. At the maintenance phase, Tom maintained 90% accuracy with slightly decreasing trend when solving word problems.

Participant 3

Sam completed none of the word problems during the baseline phase, however, the change in his performance was immediate and pronounced at the beginning of the intervention. The mean percentage of accurate responding during intervention was 86.67%. The mean level change between baseline and intervention sessions was 86.67%. Sam maintained his problem solving accuracy with a mean of 70% at a two-week follow up.

Social validity

All participants responded positively to the social validity questions during both initial conversation and at the end of the study. At the end of the study, they indicated that they enjoyed watching the video. Two of the students also indicated that they would prefer to use a tablet pc or video clips to do mathematics homework. When asked whether they would like to learn mathematics with video clips in the future, all students responded very positively. It was observed that all the students seemed excited and willing to watch the video clip each day. Both classroom teachers stated that it’s important to utilize technology. During the social validity conversation at the end of the study, teachers underlined that they enjoyed participating in the study. Each teacher reported seeing immediate progress in the students. One teacher remarked, ‘I saw that they remembered each step from day to day’. The other commented, ‘Video clip idea was really beneficial, but person hand in the video might be distraction for some students’. When they asked about their willingness for future use of the VBI, they stated that most of the students would benefit from using this type of technology in academics.

Discussion

The purpose of the current study was to examine the effects of POVM intervention via a tablet to teach mathematics word problem solving skill acquisition requiring addition with regrouping to students with ASD. Results of the study interpreted with visual analysis indicated immediate increase for every participant reaching a mean accuracy of 92.22% during intervention and 86.67% during maintenance. According to the findings, POVM intervention was effective in increasing mathematics word problem solving accuracy and skill maintenance for students with ASD.

These findings are consistent with the previous research on VBI that increased mathematics skill acquisition for students with ASD (Burton et al. 2013, Yakubova et al. 2016). However, the vast majority of research focusing on teaching academic skills consist of basic skills of reading and writing and limited research focuses on intervention and strategies to teach mathematics to students with ASD (Root 2019). Spencer and her colleagues (2014) reported that the majority of conducted research included elementary and mixed-grade levels, while very few included secondary grade level students. The results are also consistent with the findings of previous researchers related to use of handheld technology to increase learning of students with ASD (Burton et al. 2013, Cihak and Bowlin 2009, Cihak et al. 2010, Yakubova et al. 2016).

VBIs are generally accepted as evidence-based practices for students with ASD (NAC 2015), but have less commonly been systematically evaluated to improve mathematics skills acquisitions of students with ASD. The POVM intervention included elements that have well documented for sufficient scientific rigor is video modeling, explicit instruction, and use of technology (Yakubova et al. 2015). Findings of the current study extended the video modeling research literature in several ways. First, the study extended the research beyond existing literacy and non-academic areas by documenting the acquisition of mathematics skills linked to the core curriculum. Second, video modeling incorporated instructional methodologies for students with ASD who were receiving services in a private special education school increased access to the core curriculum. Third, the study extended literature including the VBI to teach mathematics skill to a new population of students, secondary grade level students in Turkey.

In addition to its support to previous VBI research, this study joined and supported research on mathematics interventions for secondary-level students with ASD. At the same time, these findings suggest that POVM instructions were easy to implement for teachers and efficient to design any subject tailored to the needs of students. Rather than providing traditional mathematics instruction in small groups, POVM creates more opportunities for teachers to provide support and feedback for each individual. Improving mathematics achievement of students with ASD is significantly important and requires a strong foundation of prerequisite skills (Yakubova et al. 2016), so broadening the scope and use of POVM according to the students’ needs and educational programs would be optimal.

The immediate increase in students’ performance might be explained with the visual support. Visually-cued instructions are an effective intervention method for individuals with ASD to acquire, generalize and maintain skills (Bellini and Akullian 2007, Bennett et al. 2017). Moreover, handheld technologies are likely to increase student’s learning by reducing extra stimuli and creating more opportunities for instruction from the first person perspective when technology is available. Computers tend to act as controlled environments with minimal distractions and attractive for the education of students with ASD (Boucenna et al. 2014). Teachers stated that being able to re-watch the video as an option appeared to be the other advantage of point-of-video modeling. It is more efficient when students can access already recorded instruction compared to waiting for the teachers’ direct assistance. Additionally, teachers can design their own videos according to students’ strengths and academics needs which will help teachers to better individualize instruction when class time is limited.

After the immediate increase in treatment phases, students’ performance continued during the maintenance phases two weeks after the completion of treatment data collection. Even if the mean percentage of problems correct in maintenance phase was lower than the treatment phase, findings showed that participants maintained their word problem solving accuracy. These individual performances showed students’ success solving addition with regrouping problems after the intervention had been removed. Moreover, these findings supported that visually cued instructions, POVM specifically, were effective for maintaining mathematics skills.

Limitations and recommendations for future research

There are at least three limitations to this investigation that are important to consider. First, limited sample size of students with ASD may not represent the general population. While the National Standards Report for evidence-based practices for ASD suggest at least three participants to reach sufficient scientific rigor (NAC 2015) and best practices in single case intervention research design suggest the replication of effect across three participants (Kratochwill et al. 2014), replication and future research should include larger sample sizes to enhance the scope of findings.

Second, this investigation included change problems as the classification of all the word problems so a natural next step would be to expand to compare problems. Efforts to address college and career readiness skills might consider POVM to teach more complex operations, especially considering that NMAP (2008) propose that all students should have access to and be successful learning content.

Third, findings from this investigation revealed a positive two-week maintenance results for participants; however, future research should include maintenance at longer intervals. In addition, collecting generalization data across settings, materials and/or persons would enhance the scientific rigor of the investigation to draw firm conclusions about the treatment effect. Acquisition and short-term maintenance of skills, which findings from this study supported, are requisite to develop long-term maintenance and application of skills in novel or practical application.

Further research is needed to employ POVM, other VBI, and the use of tablet pc with different mathematics skills including basic numeracy skills, functional mathematics, and algebra to improve mathematics knowledge of students with ASD. Sanders et al. (2005) identified rational numbers as a critical area of need for middle school students. Teachers and two of the participants stated their positive viewpoints and willingness to use VBI for future learning situations. Given the varying mathematics skills of students with ASD, VBI offers teachers a way to provide specially designed instruction specifically based on their strengths and weakness resulting in better outcomes.

Conclusion

This study examined the effects of POVM to teach mathematics word problem solving skill acquisition requiring addition with regrouping to three students with ASD in Turkey. The results from this study extend the current body of research to international settings with on a mathematics skill. Findings showed that each student demonstrated considerable improvement during intervention over baseline levels, and subsequently sustaining their performance through the maintenance phase. Although VBIs are considered as evidence-based practices in many social, communication, and growing academic areas, more research is needed in mathematics before conclusion can be generalized across participants and mathematics topics. Findings from the current study are promising as the Turkish students with ASD demonstrated acquisition and maintenance of skills to solve word problems requiring addition with regrouping after POVM instruction.

Conflict of interest

The authors report no conflict of interest.