Evaluation of Generalized Performance Across Materials When Using Video Technology by Students With Autism Spectrum Disorder and Moderate Intellectual Disability

Abstract

The purpose of this study was to evaluate the ability of four high school–aged students with a diagnosis of autism spectrum disorder and moderate intellectual disability to generalize performance of skills when using materials different from those presented through video models. An adapted alternating treatments design was used to evaluate student performance when using materials depicted directly in the custom-made video programs compared with materials different from those depicted in the custom-made video programs. Results indicated that overall, students committed more errors and required more sessions to reach criteria when using the generalized materials; however, three of the four students completed tasks under Comparisons 1 and 2 with 100% accuracy when using materials that differed from those in the video model.

Keywords

video modeling generalization autism spectrum disorders moderate intellectual disability

The purpose of this study was to evaluate generalized performance of four high school–aged students with a diagnosis of autism spectrum disorder (ASD) and moderate intellectual disability when using materials different from those presented through video models. An adapted alternating treatments design was used to evaluate student performance when using materials depicted directly in the custom-made video programs compared with materials different from those depicted in the custom-made video programs. Results indicated that overall students committed more errors and required more sessions to criteria when using the generalized materials; however, three of the four students completed tasks under Comparisons 1 and 2 with 100% accuracy when using materials that differed from those in the video model.

Generalization of skills across materials and settings has long been a concern for professionals responsible for the programming and planning of instruction for persons with intellectual disabilities (ID), including those with ASD (Stokes & Baer, 1977). It is recognized that generalization cannot automatically be expected to occur without proper planning and careful selection of teaching materials and procedures. Video-based instruction (VBI) has been researched as a tool for teaching students with ASD and as a tool for promoting generalization of skills (Charlop-Christy & Daneshvar, 2003; Charlop-Christy, Loc, & Freeman, 2000; Delano, 2007; LeBlanc et al., 2003; Sherer et al., 2001). VBI provides the opportunity to present multiple examples of materials (i.e., food labels; Mechling, Gast, & Krupa, 2007), settings (Mechling & Gast, 2003), and multiple models of behaviors (Mechling & Hunnicutt, 2011). Although the use of multiple examples can be important for teaching some skills, it is also important to recognize the time and effort involved in creating effective video models (LeBlanc, 2010). These efforts and demands on time will increase if multiple, custom-made video programs, to meet the individual needs of several students within a classroom or training facility, are needed. Teachers express concern for the time needed to create custom-made videos (Mechling, Ayres, Foster, & Bryant, 2013), and therefore, practitioners may be reluctant to create multiple video programs that represent personalized materials rather than one video program that can be used across students, environments, and materials. For example, an interventionist providing support to persons living in semi-independent living environments may be responsible for developing video cooking recipes or laundry video training programs for several clients. In such a scenario, it would be more convenient and efficient if one generic video could be used across students or clients. The question is whether users with ID and ASD can perform skills using materials in their personal settings (i.e., kitchen) that differ from those represented in the generic video.

The effectiveness of commercially available, pre-made, instructional videos is still under scrutiny with only a limited number of studies being conducted (Allen, Wallace, Renes, Bowen, & Burke, 2010; Ayres & Cihak, 2010; Ayres, Maguire, & McClimon, 2009; Blum-Dimaya, Reeve, Reeve, & Hoch, 2010; Mechling et al., 2013; Palechka & MacDonald, 2010; Rosenberg, Schwartz, & Davis, 2010). Although convenient, these commercial products may not be effective if the developers are inexperienced in working with persons with ID and ASD and fail to attend to the detail that a behavior analyst or special educator, familiar with teaching persons with ID and ASD, may understand (LeBlanc, 2010). In a study comparing custom-made and commercially available video programs, Mechling et al. (2013) summarized that student errors, when using the commercially available videos, were influenced by the inability of students to understand concepts that were inadequately presented on the commercial videos (i.e., waiting for ingredients to boil; the act of measuring ingredients).

Although preliminary research has addressed the concern for using commercially produced video programs among students with ID and ASD, no research to date has evaluated the ability of students with ASD and moderate intellectual disability to generalize performance across materials and equipment when custom-made video programs are used. Mechling et al. (2013) recommended that research be conducted that examines generalization across materials and equipment not depicted in custom-made videos. The purpose of the present study was to follow this recommendation and to compare the task completion of students with ASD and moderate ID when using (a) materials depicted directly in custom-made video programs and (b) materials different from those depicted in custom-made video programs.

Method

Participants

Four males who attended a public high school participated in the study and each met the school systems’ requirements for services under the ASD moderate intellectual disability criteria. Washington and Michael were in the same classroom whereas Chase and Aden were in a classroom in the same wing of the school. Each classroom was designed to serve students with moderate intellectual disability. The four students also participated in a previous study comparing commercially available and instructor-created video models (Mechling et al., 2013). Students were selected based on their ability to follow a video model and to imitate behaviors, Individualized Education Plan (IEP) objectives to complete home living skills, classification and sorting skills, and attention to a task for 15 min. The participants were the only students within the school setting to have a diagnosis of ASD and meet the criteria stated above. Table 1 provides information about each student.

Table 1.

Student Characteristics.

Student	Age	Gender	Diagnoses	IQ	Adaptive Behavior
Washington	17 years 1 month	M	ASD	Very low^a	51^c
			Moderate ID	Below 50^b
Chance	18 years 6 months	M	Fragile X syndrome	45^d	69^c
			Moderate ID
Aden	19 years 5 months	M	ASD	55^e	48^c
			Moderate ID
Michael	15 years 10 months	M	ASD	48^f	55^c
			Moderate ID

Woodcock–Johnson Test of Cognitive Abilities–Third Edition. ^bBayley Scales of Infant Development–Second Edition. ^cVineland Adaptive Behavior Scale. ^dWechsler Preschool and Primary Scale of Intelligence–Revised. ^eKaufman Assessment Battery for Children. ^fStanford–Binet Intelligence Scales–5th Edition.

Setting, Equipment, Materials, and General Procedures

All baseline and comparison sessions were implemented in the home living room of the high school that was used to teach functional daily living skills to students enrolled in special education classrooms. The room contained a complete kitchen with a stove, refrigerator, sink, microwave, banquet-size kitchen table, and a stack washer and dryer. Students were brought to the room individually and no other students were present in the home living room unless someone came in to put something in the washer/dryer and to heat food in the microwave.

All custom-made videos were made using a SONY digital video camera and edited using Windows Movie Maker. Videos were played on a Dell Latitude E6410 computer with video clips inserted and played using PowerPoint slides. Two different types of tasks were selected for the comparison study: recycling and loading a dishwasher. For each type of task, there were two different sets (i.e., recycling, Set 1: plastic, glass, aluminum; Set 2: paperboard, cardboard, paper). Within each set, three functionally independent groups of materials were selected to present a different set and group for the actual materials (materials used by students were identical to those in the video) and generalized materials (materials used by students were different from those in the video) to each pair of students during the comparison conditions (see Table 2).The third group of materials within each set served as control materials (task performed without a video instruction).

Table 2.

Actual and Generalized Materials Used Across Tasks and Sets.

Task: Recycling
Set 1		Set 2
Group A	Group B	Group A	Group B	Control set
Actual materials	Actual materials	Actual materials	Actual materials
	Washington/Chance		Washington/Chance
	Comparison 2		Comparison 1
	Michael/Aden		Michael/Aden
	Comparison 1		Comparison 2
Generalized materials	Generalized materials	Generalized materials	Generalized materials
Washington/Chance		Washington/Chance
Comparison 1		Comparison 2
Michael/Aden		Michael/Aden
Comparison 2		Comparison 1
Plastic containers	Plastic containers	Paperboard containers	Paperboard containers	Paper with fasteners
Moisturizer	Vitamins	Peanut butter cracker box	Granola bars	Large clasp envelope
Lean Cuisine tray	Peroxide	Frozen artichokes	Lean Cuisine box	Small clasp envelope
Shampoo	Mustard	Muffin mix	Pop Tarts	Green clasp envelope
Coke	Liquid soap	Saltine crackers	Spaghetti box	File folder
Blueberries	Carrot/celery	Butter box	Dried soup	Notebook
Glass containers	Glass containers	Cardboard containers	Cardboard containers	Cardstock
Olives	Capers	Red tube	Mail tube	File folder
Salsa	Dressing	Cardboard insert	Router box	Brown folder
Relish	Pickles	Flat box	Cardboard box	Envelope
Teriyaki	Worcestershire sauce	Electric saw box	Flat tube	Business cards
Food coloring	Vinegar	White shipping box	Box lid	Construction paper
Aluminum containers	Aluminum containers	Paper	Paper	Paper
Foil gift bag	Aluminum foil ball	Cooking Light magazine	Eddie Bauer magazine	Notebook paper
Tuna fish	Cat food	Bed Bath & Beyond ad	New River Pottery ad	Graph paper
Diet Coke	Pepsi	Christmas card	Thanksgiving card	Blue paper
Peas/carrots	Corn	Replacements envelope	Electric bill envelope	Yellow paper
Green beans	Chicken broth	Barnes & Noble ad	Forever 21 ad	Scrap paper
Task: Loading dishwasher
Set 1		Set 2
Group A	Group B	Group A	Group B	Control Set
Actual materials	Actual materials	Actual materials	Actual materials
Washington/Chance	Washington/Chance
Comparison 2	Comparison 1
Michael/Aden	Michael/Aden
Comparison 1	Comparison 2
Generalized materials	Generalized materials	Generalized materials	Generalized materials
	Washington/Chance	Washington/Chance
	Comparison 1	Comparison 2
	Michael/Aden	Michael/Aden	Comparison 2	Comparison 1
Muffin pan	Acorn pan	Pink pattern plate	Green pattern plate	Spatula
Rinse front	Rinse front	Rinse front	Rinse front	Rinse front
Rinse back	Rinse back	Rinse back	Rinse back	Rinse back
Put in	Put in	Put in	Put in	Put in
Plastic cutting board	Wooden cutting board	Pink pattern saucer	Green pattern saucer	Measuring cup
Rinse front	Rinse front	Rinse front	Rinse front	Rinse inside
Rinse back	Rinse back	Rinse back	Rinse back	Rinse outside
Put in	Put in	Put in	Put in	Put in
Small pan lid	Large pan lid	Clear bowl	Green pattern bowl	Funnel
Rinse inside	Rinse inside	Rinse inside	Rinse inside	Rinse inside
Rinse outside	Rinse outside	Rinse outside	Rinse outside	Rinse outside
Put in	Put in	Put in	Put in	Put in
Small pan	Large pan	Clear glass	Pattern glass	Sifter
Rinse inside	Rinse inside	Rinse inside	Rinse inside	Rinse inside
Rinse outside	Rinse outside	Rinse outside	Rinse outside	Rinse outside
Put in	Put in	Put in	Put in	Put in
Small red skillet	Large gray skillet	Gray coffee cup	Green coffee cup	Wooden spoon
Rinse inside	Rinse inside	Rinse inside	Rinse inside	Rinse front
Rinse outside	Rinse outside	Rinse outside	Rinse outside	Rinse back
Put in	Put in	Put in	Put in	Put in

Within Set 1 of the recycling task, there were two groups of materials (Groups A and B) and each group contained three different types of materials to recycle: glass, plastic, and aluminum. Within each type of material, there were 5 different items for a total of 15 items to recycle within each set and group. Within Set 2 of the recycling task, there were two groups of materials (Groups A and B) and each group contained three types of materials to recycle: paperboard, cardboard, and paper. Within each type of material, there were 5 different items for a total of 15 items to recycle within each set and group. For the control set, there were three types of materials (paper with fasteners, cardstock, and paper) and 5 items within each type of material for a total of 15 items to recycle (see Table 2). When sorting each set, students sorted items into three separate containers: large box, brown grocery bag, and a white shopping bag. The video models were separated into three PowerPoint slides (i.e., one for plastic, one for glass, and one for aluminum). Each video model demonstrated sorting all five items into the correct container (i.e., glass into the white bag). Students could watch the entire video model followed by placing the 5 items into the correct container or place the items into the correct container as the video played (simultaneous video modeling; Sancho, T. M. Sidener, Reeve, & Sidener, 2010). The instructor then proceeded to activate the next PowerPoint slide with the video model of the next 5 items to be sorted (i.e., aluminum into the brown bag).

Similar to the number of sets and groups used for recycling, the task of loading the dishwasher contained Sets 1 and 2 with two groups per set (Groups A and B) with five different items per group to rinse and place in the dishwasher (see Table 2). Each of the five PowerPoint slides contained a separate video demonstrating rinsing two sides of each item and placing the item in a specific location (i.e., top left rack) in the dishwasher. Students could watch the video and then complete the cluster of three steps using a video modeling procedure or complete the steps as the video played (simultaneous video modeling).

For each task (recycling and loading the dishwasher), students were assigned to different groups within each set prior to the start of the study and groups were counterbalanced across pairs of students and comparison conditions (see Tables 2 and 3). For example, during the Comparison 1 condition when recycling, Washington and Chance recycled the actual materials from Set 2 (paperboard, cardboard, paper) and Group B (i.e., paperboard: granola bars, Lean Cuisine box, Pop Tarts, spaghetti box, dried soup) that they saw on the video model. When recycling the generalized materials, they recycled Set 1 (plastic, glass, aluminum) and Group A (i.e., plastic containers: moisturizer, Lean Cuisine tray, shampoo, Coke, and blueberries) while watching the video with the other materials from Set 1, Group B (i.e., vitamins, peroxide, mustard, liquid soap, carrot/celery container) being sorted into the correct bins. During the Comparison 2 condition, Washington and Chance recycled the actual materials from Set 1, Group B, and generalized materials from Set 2, Group A.

Table 3.

Counterbalancing of Sets Across Comparison 1 and Comparison 2 Conditions Using Actual and Generalized Materials.

	Comparison 1		Comparison 2
Student	Actual materials	Generalized materials	Actual materials	Generalized materials
Michael	Set 1 Group B	Set 2 Group A	Set 2 Group B	Set 1 Group A
Aden	Set 1 Group B	Set 2 Group A	Set 2 Group B	Set 1 Group A
Washington	Set 2 Group B	Set 1 Group A	Set 1 Group B	Set 2 Group A
Chance	Set 2 Group B	Set 1 Group A	Set 1 Group B	Set 2 Group A

Data Collection and Response Definitions

The number of correct and incorrect student responses for sorting each item of the recycling tasks and each step for loading the dishwasher were individually recorded during each session of the study. A correct response was defined as initiating a step within 3 s of completion of the video model or previous step and completion of the step within 10 s. During baseline, the first step of each task was recorded as correct if it occurred within 3 s of the general task direction provided by the instructor. Students could also self-correct, meaning they could remove an item from a container during the recycling task or reposition an item in the dishwasher if they did so within 10 s of initiation of the step. Three types of incorrect responses were possible. A latency error was recorded if the student did not initiate the step within 3 s, a duration error was recorded if the student did not complete the step within 10 s, and a topographic error was recorded if the student completed the step incorrectly. When sorting items for the recycling tasks, the sequence of steps was not critical and the step was recorded as correct as long as the item was placed in the correct container. When loading the dishwasher, students were required to rinse both sides of the item, but the sequence was not critical and the steps were recorded as correct as long as both sides were rinsed. For loading the dishwasher, students were required to select an item to match the video model and place the item in the exact location shown on the video for the step to be counted as correct. If placed in an incorrect location, the response was scored as a topographical error. The instructor provided no assistance with task completion or correction of errors.

Experimental Design

An adapted alternating treatments design (Wolery, Gast, & Hammond, 2010) was used to evaluate students’ completion of tasks when using materials identical to those depicted in video models (actual materials) and when using materials different from those depicted in video models (generalized materials). The study included three conditions: baseline, Comparison 1, and Comparison 2. The baseline condition evaluated students’ abilities to complete the recycling and loading the dishwasher tasks prior to the introduction of video and to measure equivalence of performance across sets regardless of the materials used (Sindelar, Rosenberg, & Wilson, 1985). The Comparison 1 condition compared students’ abilities to complete tasks using actual materials (materials identical to those used in the video model) and generalized materials (materials different from those used in the video model). During this comparison condition, students used one set and group of actual materials (i.e., Set 1, Group B, of glassware, plastic, and aluminum) and a different set and group of generalized materials (i.e., Set 2, Group A, of paperboard, cardboard, and paper). Different sets of materials were used within a condition to diminish the likelihood of a carryover effect.

The Comparison 2 condition then evaluated students’ abilities to generalize the use of materials following performance with the actual materials of that set during Comparison 1. In other words, Washington and Chance used the actual materials of Set 2 (Group B) during Comparison 1, and during Comparison 2, they used generalized materials (materials not in the video) of Set 2 (Group A) while watching the video with the actual materials (Set 2, Group B). The Comparison 2 condition also measured any change in student performance when each watched and performed the task with the actual materials from Condition 1 rather than the video with the generalized materials. This served to evaluate any differences in performance that may have been affected by material types rather than the use of actual and generalized materials. Actual and generalized video models were counterbalanced across tasks and students to also control for task difficulty. Table 3 and the top portion of each task description in Table 2 show the counterbalancing across sets and groups for Conditions 1 and 2.

Students completed tasks with materials from the control set without the use of video models during the baseline and Comparison 1 conditions. During the Comparison 2 condition, students completed the task using video models with actual materials depicted on the video. Extension of the baseline condition, using the control set in the first comparison condition and application of video modeling in the second comparison condition, served to evaluate possible multiple treatment interference (threat to internal validity) and the effects of history and maturation, and to allow intra-subject replication.

Two types of tasks, recycling and loading a dishwasher, were used to replicate the evaluation across tasks. The two tasks were introduced and implemented simultaneously. Sessions were conducted individually across all conditions and the two tasks were implemented during each session. Both video models were evaluated (actual and generalized) for each task during each session. Counterbalancing was provided across sessions between the actual video models and generalized video models for each task to minimize the possibility of sequencing effects. The two types of tasks were selected due to their functional use in school (schoolwide recycling program) and at home (washing dishes during meal preparation). Both tasks included placing items into specific locations with a given visual (video) cue.

Procedures

Baseline procedures

During baseline, students were assessed individually on completion of the recycling and loading the dishwasher tasks using their assigned group of materials and the control set of materials without the use of video models. For the recycling sessions, 15 items were intermixed and placed randomly on the banquet table in front of the three containers (15 trials). For loading the dishwasher, the dishwasher door was opened and the 15 items were randomly placed on the counter above the dishwasher (15 trials). Students could sit or stand when completing the recycling task and stood when loading the dishwasher. The instructor provided a general task direction, “Sort the containers for recycling” or “Load the dishwasher” at the beginning of the session. Students were given 3 s to initiate sorting of an item or rinsing and placing an item in the dishwasher and 10 s to complete the step. No prompts were provided by the instructor. Errors were ignored and verbal praise was provided for correct responses and for general attending to the task. At the end of the session, students were reinforced with a choice of snack items. Baseline continued for a minimum of three sessions and until data were stabilized for all sets within a task with no improvement in performance.

Comparison 1 procedures

Procedures for the Comparison 1 condition were identical to those used during baseline except that students used video models to prompt the completion of tasks. The laptop computer was placed on the left end of the banquet-size table for the recycling task and on the counter above the dishwasher for loading the dishwasher. The instructor sat in front and to the left of the laptop computer during the recycling tasks and stood to the right of the laptop computer during the loading of the dishwasher task.

For recycling, students watched a video model demonstrating sorting items into the correct container (i.e., glass into the white bag). Students could watch the entire video model followed by placing the five items into the correct container or place the items into the correct container as the video played. Students could also request that a video segment be re-played. The instructor advanced the program to the subsequent PowerPoint slide when the student stopped placing materials into a container. The instructor proceeded to activate the next PowerPoint slide with the next video model (i.e., aluminum into the brown bag) until all three video models were played.

When loading the dishwasher, each PowerPoint slide contained a separate video demonstrating rinsing two sides of each item and placing the item in a specific location (i.e., silverware basket) in the dishwasher. Students watched a video and then completed the cluster of three steps (rinse the front, rinse the back, place item in the dishwasher). Students could also request that a video segment be re-played. The instructor then advanced the program to the next slide which played the video prompt. This continued until the five video models were played.

Table 3 and the top portion of each task description in Table 2 show the groups of materials within each set used by the pairs of students. When recycling or loading the dishwasher, if a student used the actual materials depicted on the video model from Set 1, then he used generalized materials not depicted on the video model from Set 2 during the first comparison condition.

The control set of materials was intermittently evaluated on an average of every three to four sessions during the Comparison 1 condition without the use of a video model. The Comparison 1 condition was implemented for a minimum of six sessions. If one video intervention reached criteria (100% correct responding for one session), both video interventions were continued for twice the number of sessions it took the first video intervention to reach criteria and until there was no improvement in performance across both video interventions (Wolery et al., 2010). In addition, if both video interventions reached 100% correct responding, the condition was terminated after a minimum of six sessions.

Comparison 2 procedures

Procedures for the Comparison 2 condition were identical to those of the Comparison 1 condition with the exception of the materials being sorted or loaded into the dishwasher and evaluation of the control set of materials using a video model. Table 3 and the top portion of each task description in Table 2 show the groups of materials within each set used by the pairs of students during the Comparison 2 condition. When recycling or loading the dishwasher, if a student used the actual materials depicted on the video model from Set 1 during the Comparison 1 condition, then he used actual materials from Set 2 during the Comparison 2 condition. Likewise, if he used generalized materials (not depicted on the video model) from Set 2 during the first comparison condition, he used generalized materials from Set 1 during the Comparison 2 condition. The Comparison 2 condition was implemented for a minimum of six sessions and followed the same guidelines for criterion levels and termination as those for the Comparison 1 condition.

The evaluation of the control set of materials for recycling and loading the dishwasher used video models depicting the actual materials of the control set. These sessions were implemented using procedures identical to those in Conditions 1 and 2 when students completed sets with the actual materials depicted in the video models.

Social Validity

Two classroom teachers and two teacher assistants, who taught the students on a daily basis and who were not involved in the study, ranked the following statements from 1 (strongly agree) to 5 (strongly disagree) at the conclusion of the study:

Custom-made videos can be an effective tool for teaching home living skills to students with ASD and moderate intellectual disability;

Creating my own videos for teaching skills has the potential to help with the generalization of skills to new materials and settings; and

If taught to do so, I would create my own video programs for teaching home living skills to students with ASD and moderate intellectual disability.

The two teacher assistants were included because they were also responsible for creating instructional materials used by students in the classroom.

Reliability

Data were recorded for interobserver agreement (IOA) by one of two independent research assistants who also simultaneously scored procedural reliability data. The research assistants were trained on data collection procedures prior to a series of research studies for which they were involved. For the recycling task, independent observation of the behavior of the instructor and students was conducted for 48% of all sessions across the three conditions and for 33% of the baseline condition and 50% of the Comparison 1 and Comparison 2 conditions. For Michael, data were collected during 38.9% of his sessions; for Aden, 42.8% of his sessions; for Chance, 56.8% of his sessions; and for Washington, 45.8% of his sessions.

For loading the dishwasher, independent observation of the behavior of the instructor and students was conducted for 55.6% of all sessions across the three conditions and for 29.4% of the baseline condition and 57.8% of the Comparison 1 and Comparison 2 conditions. For Michael, data were collected during 63.2% of his sessions; for Aden, 55% of his sessions; for Chance, 50% of his sessions; and for Washington, 54.5% of his sessions.

The mean IOA was calculated by dividing the number of agreements on correct responses by the number of agreements plus disagreements and multiplying by 100. The IOA for the recycling task was 99.5% across students and conditions. For Michael and Aden, the mean IOA was 100% across conditions. For Chance, the mean IOA was 99.4% with a range of 86.7% to 100%, and for Washington, the mean IOA was 98.9% with a range of 93.3% to 100%. The evaluators were able to look into the containers after each student sorted the items that contributed to the high level of agreement.

The IOA for loading the dishwasher was 98.4% across students and conditions. For Michael, the mean IOA was 98.4% with a range of 80% to 100%. For Aden, the mean IOA was 98.4% with a range of 86.7% to 100%. For Chance, the IOA was 98.7% with a range of 93.3%–100%, and for Washington, the IOA was 98.2% with a range of 86.7% to 100%. When loading the dishwasher, students quickly completed the rinsing steps of the task (i.e., front and back of the spatula) and disagreement occurred for these steps.

The percentage of correctly performed behaviors by the instructor (procedural fidelity) was calculated by summing the total number of correctly performed steps and dividing that number by the total possible behaviors across the baseline and comparison conditions. The targeted behaviors included (a) delivering the general task directions, (b) presenting the correct video model in a counterbalanced approach, (c) placing materials on the table or counter, (d) correct playing and advancing of the video models, (e) waiting 3 s for students to initiate task steps, and (f) waiting 10 s for students to complete task steps. The mean percentage of correct responses for the instructor during recycling was 99.7% (range = 98.9%–100%). The majority of the errors occurred when materials were incorrectly placed on the table (i.e., spaghetti box placed on the table with Group A). Other errors occurred when the containers were placed in the incorrect order on the table, the instructor started an incorrect video during the start of one session, and the video skipped ahead during another session. The mean percentage of correct responses for the instructor during loading of the dishwasher was 99.8% (range = 99.5%–100%). Similar to the recycling task, the majority of errors were committed when incorrect materials were placed on the counter or were missing. In addition, Aden broke a glass during one session and this was recorded as a procedural error as the glass was unavailable to rinse and place in the dishwasher.

Results

Figures 1 to 4 illustrate each student’s correct performance when completing the recycling and loading the dishwasher tasks across actual and generalized materials. The results indicate that overall students performed somewhat better during Condition 1 when using materials that were the actual materials depicted in the video models compared with the use of materials (generalized) different from those in the video models. However, results also indicate that the four students were able to generalize performance of the tasks across materials that differed from those they saw on the video models.

Figure 1.

Percentage of steps correctly completed by Michael during baseline and comparison conditions using actual materials (closed circles), generalized materials (open squares), and the control set (open triangles) across Set 1 (recycling) and Set 2 (loading the dishwasher).

Figure 2.

Percentage of steps correctly completed by Aden during baseline and comparison conditions using actual materials (closed circles), generalized materials (open squares), and the control set (open triangles) across Set 1 (recycling) and Set 2 (loading the dishwasher).

Figure 3.

Percentage of steps correctly completed by Washington during baseline and comparison conditions using actual materials (closed circles), generalized materials (open squares), and the control set (open triangles) across Set 1 (recycling) and Set 2 (loading the dishwasher).

Figure 4.

Percentage of steps correctly completed by Chance during baseline and comparison conditions using actual materials (closed circles), generalized materials (open squares), and the control set (open triangles) across Set 1 (recycling) and Set 2 (loading the dishwasher).

Recycling

During the baseline condition, students placed some of the materials into the correct containers in a random fashion. Michael and Aden placed all of the materials into one container and therefore correctly placed 5 of the 15 items correctly (33.3%) by default. During the Comparison 1 condition, two of the four students reached criterion levels with both actual and generalized materials. Washington was the only student who was unable to recycle at the criterion level using the generalized materials and Aden was unable to recycle at the criterion level using the actual materials. Washington clearly performed better when using the actual materials shown in the video model, but he reached 86.7% correct performance in the fifth session when using materials that differed from those in the video model. Likewise, Chance showed a clear difference in the data paths with only one overlapping data point in Session 3; however, his performance continued to improve with both sets and therefore the condition was extended beyond the six-session minimum and he reached criteria with both video models in Session 10. Examination of the groups of materials and errors made by students showed no consistent pattern with the exception of Chance who confused plastic in the generalized materials and put the plastic items with glass. Washington’s errors with the generalized materials were spread across aluminum and plastic, but he consistently placed the glass in the correct containers.

Interestingly, Michael and Aden both reached criterion levels earlier with the generalized materials (paperboard, cardboard, paper) and Aden was unable to perform the recycling task above 93.3% correct when using the actual materials (glass, plastic, aluminum) and had difficulty placing the Pepsi can and cat food can with the aluminum products. This may have indicated a difference in the level of difficulty across material types (i.e., discriminating between glass, plastic, and aluminum verses paperboard, cardboard, and paper).

During Comparison 2, Aden reached criteria with both the generalized and actual materials while Michael’s performance with the generalized set remained at 93.3% because he continued to place the foil gift bag in the wrong container. Chance also reached criterion level with both actual and generalized materials, but he performed 100% of the steps correctly when using the actual materials (glass, plastic, aluminum) in Session 4 and performed better with the actual materials during 75% of the sessions. His errors were inconsistent across the two sets. Washington reached the criterion level with both actual and generalized materials, but other than Session 6 when he performed 100% correctly with both sets of materials, he consistently performed better when using the generalized (paperboard, cardboard, paper) materials and had difficulty sorting the plastic containers.

Loading the Dishwasher

With the exception of Aden, students recognized the need to rinse the items during the baseline condition, but did not always rinse both sides. In addition, they placed everything in either the top or bottom rack of the dishwasher and therefore were given credit for some items by default. Change in performance was abrupt and noticeable for Michael and Aden when the video models were introduced in Comparison 1. Both students reached criterion with the actual and generalized sets of materials, but overall performed better with the actual materials.

With the exception of one session, Washington’s change in the level of performance was also notable for both the actual and generalized materials with the introduction of the video models during Comparison 1. He also performed more steps correctly when using the actual materials with the exception of Session 3 and the final session. He reached criterion with the actual materials in Session 6 and therefore Comparison 1 had to be conducted for 12 sessions; however, he performed 100% of the steps correctly in Session 11.

Chance demonstrated overlapping data on 57% of the sessions and reached criterion with the actual and generalized materials in Session 7. In comparison with the generalized materials, there were three sessions in which he performed more steps correctly with the actual materials.

During Comparison 2, three of the four students demonstrated the generalized use of materials following the use of the actual materials depicted on the video models in Comparison 1. Chance was the only student who demonstrated difficulty generalizing across materials, but his data also fluctuated when using the actual materials, indicating that the level of task difficulty or some other factor (i.e., distractibility) may have influenced his performance. He did perform 100% of the steps when using the actual materials (Session 4) and generalized materials (Session 7) during the final condition.

Control Set

Performance levels remained low for all students when performing the recycling and loading the dishwasher tasks without the use of video models during the baseline and Comparison 1 conditions. Chance demonstrated an increased level in loading items into the dishwasher during the final sessions of Comparison 1, but he also demonstrated a noticeable change in his level of performance when the video model was introduced in the Comparison 2 condition. Three of the four students reached 100% correct performance when the video model was applied to the control set for recycling actual materials and all students reached criterion-level performance when loading the dishwasher with the video model and control set of actual materials. Results discredit the likelihood of multiple treatment interference and the effects of history and maturation and also provide intra-subject support for the use of video modeling with actual materials.

Error Analysis

Across Comparisons 1 and 2, students performed more errors when using generalized materials (65.5%) compared with actual materials (34.5%). During Comparison 1, there was a noticeable difference in errors when using the generalized materials (71.8%) compared with the actual materials (28.2%); however, during Comparison 2, there was a minimal difference in the percentage of errors across materials and more errors were committed when using the actual materials (53.5%) compared with the generalized materials (46.5%). Differences in performance were more evident when using generalized recycling materials (76% errors) compared with actual recycling materials (24% errors) during Comparison 1. This visible difference in performance did not occur during Comparison 2 when students committed nearly 50% of the errors regardless of whether generalized (51.6% errors) or actual (48.4% errors) recycling materials were used. When loading the dishwasher during Comparison 1, students committed 66.4% of the errors when using generalized materials compared with 33.6% of the errors using actual materials. This difference did not hold during Comparison 2 when students committed more errors (61.5%) when using the actual dishes compared with generalized dishes (38.5% errors).

Further examination of errors indicates that students demonstrated more difficulty when sorting the recycling materials (57.4% errors) compared with placing the dishes in the dishwasher (42.6% errors) regardless of whether actual or generalized materials were used. The difference was most noticeable during Comparison 1 when using generalized recycling materials. Students committed almost half of the errors (42.6%) when using generalized recycling materials compared with 29.2% when using generalized dishes for loading the dishwasher and 14.8% when using actual dishes and 13.4% when using actual recycling materials. All errors across conditions and students were topographical in nature.

Social Validity

The four teachers reported (M = 1.25) that they considered custom-made videos to be effective tools for teaching home living skills to their students and that creating these videos (M = 1.25) had the potential for promoting generalization of skills to new materials and settings. Two of the teachers strongly agreed that they would create their own video programs if taught to do so and the remaining two teachers agreed with this statement. One teacher wrote on the form that she would be willing to create her own videos, “If provided with the equipment and supplies.”

Discussion

Results of this study have implications when considering the use of generic video-based programs across multiple students and settings. As one might expect, students did commit more errors when using generalized materials not depicted in the video models and required more sessions to reach criterion levels of performance. However, during Condition 1, with the exception of one student (Washington) and one task (recycling), students performed all tasks at criterion levels (100% correct) when using generalized materials. During this first condition, students were required to perform the tasks of recycling and loading the dishwasher using materials that differed from those on the video without any previous experience using the actual materials. For the purpose of the research study, correct performance by students was accomplished with no provision of additional support by the instructor. Results are promising for using generic videos with multiple students even if the materials may differ from those they see in the video. In applied settings, instructors should also consider the use of other prompting strategies to increase the efficiency of using VBI with materials that differ from those in the original video. Although important for experimental control purposes in this study, it is recommended that video modeling be combined with other instructional strategies to increase the likelihood that efficient learning will take place (Goldsmith & LeBlanc, 2004).

Correct performance increased when using generalized materials in Condition 2 in comparison with Condition 1. Data from the second condition were supportive of students’ abilities to generalize performance across new materials following the use of the actual materials depicted in the video. With the exception of Michael when performing the recycling tasks, students performed all tasks 100% correctly when using both actual and generalized materials and did so with fewer sessions to reach criteria than demonstrated in Condition 1. These results indicate that students were able to generalize use of materials following traditional methods of instruction in which they were taught with one set of materials and then required to perform (generalize) with another set. In contrast, during Condition 1, they demonstrated more difficulty when required to initially use materials that differed from those in the video.

Results also indicate that types of materials may affect students’ use of materials across generic video models. More errors were committed when recycling compared with loading the dishwasher regardless of whether the materials used were those actually depicted in the video models or different from those in the video models. For example, it is likely that even if two plates or two coffee cups differ in color or pattern from those depicted in a video, their similarities will be strong enough to allow for generalized performance. In contrast, a cardboard box and a cardboard tube may lack enough similarities to allow for such generalized performance. Persons developing video models to be used across students, settings, and materials should attend to the principals of general case programming (Horner, Sprague, & Wilcox, 1982; Westling & Fox, 2009) whereby close examination of the range of examples is made prior to the selection of the materials, and target materials to be used in the video models are clear and best examples from that range and not just those convenient or available to the video producer.

Furthermore, the type of material and discrimination requirements affected results of the recycling tasks. More errors were committed with Set 1 when sorting plastic containers (51.4%) compared with aluminum (35.6%) and glass (13%) containers regardless of whether the materials were actual or generalized. However, these errors were even more pronounced when using generalized materials. Sorting generalized plastic, aluminum, and glass containers contributed to 76.3% of the errors compared with 23.7% of the errors with Set 1 when using actual materials. In contrast, these differences between actual and generalized materials were not as noticeable when sorting Set 2: actual materials, 51.8% errors; generalized materials, 48.2% errors.

Although the focus of the study was not on the use of simultaneous video modeling compared with video modeling, due to the short length of each video, students were permitted to watch the entire video model followed by completing the step or to place items into the correct location as the video played. All students elected to use simultaneous video modeling when completing the recycling task. No consistent behaviors for using the video were shown when loading the dishwasher although anecdotally it was noted that students began to place items in sync with the video as they became more familiar with the tasks (e.g., later sessions). Future research endeavors should target the measurement of student responses across tasks when using video modeling and simultaneous video modeling. Taber-Doughty, Patton, and Brennan (2008) found both delayed video modeling and simultaneous video modeling effective in their comparison study; however, the stated preferred system by each student was more effective for two of the three participants. Of further interest is whether students will perform better if they are provided with an opportunity to choose the video modeling procedure that best fits their individual learning styles.

A recognized limitation of the current study is that the design excludes information regarding how students would perform if another component (i.e., instructor prompts) was provided when using the generalized materials. It is likely that students would have performed tasks more accurately if shown the similarities between the actual materials depicted in the video models and those that they were using (generalized materials). It is recommended that this strategy be used in future research or when teaching students within an applied setting.

A further limitation is that only two sets of tasks were evaluated (recycling and loading a dishwasher). Future research should evaluate students’ use of generic video models across additional tasks and types of tasks (e.g., chained responses) and whether this behavior extends beyond students diagnosed with a moderate intellectual disability and ASD.

While the current study sought to evaluate custom-made video models, results have implications for the use and design of commercial products as well custom-made products. Failure of students to perform tasks correctly when using commercial video products may be due to concepts being inadequately presented on the video (Mechling et al., 2013), rather than their ability to generalize performance across materials different from those on the video, or due to products being created by persons inexperienced in working with persons with ID and ASD (LeBlanc, 2010). Results of this study and the study by Mechling et al. collaborate the statement by Ramdoss et al. (2012) who suggested that the quality of instructional materials and effective interventions, when using computer-based interventions, is dependent on the quality of the entire presentation and knowledge of the developers of the curriculum when using general purpose software programs such as PowerPoint and Windows Movie Maker than on the software itself. Results of the current study demonstrate students’ abilities to complete tasks even when video models use materials different from those in their environments if close attention is paid to the selection and presentation of materials used in the video models. However, the study did not include measures of generalization across settings whereby equipment such as the dishwasher, recycling bins, and sink may differ from those in the video model. Future research should attend to these differences as possible contributors to generalization by users. Each of these factors has implications for the growing use of “apps,” the expressed need for publicly available or purchasable video tools (Goldsmith & LeBlanc, 2004), and the vital need for clinicians and researchers, knowledgeable in working with persons with ID and ASD, to work with programmers and engineers to develop and modify technology-based interventions (Goldsmith & LeBlanc, 2004).

Footnotes

Authors’ Note

The opinions expressed are those of the authors and do not represent views of the Institute of Education Sciences or the U.S. Department of Education.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported here was supported by the Institute of Education Sciences, U.S. Department of Education, through Grant R324A100094 to the University of Georgia.

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.

Ayres

K. M.

Cihak

D. F.

(2010). Computer and video based simulation of food preparation skills: Acquisition, generalization and maintenance. Intellectual and Developmental Disabilities, 48, 195–208.

Ayres

K. M.

Maguire

McClimon

(2009). Acquisition and generalization of chained tasks taught with computer-based video instruction to children with autism. Education and Training in Developmental Disabilities, 44, 493–508.

Blum-Dimaya

Reeve

S. A.

Reeve

K. F.

Hoch

(2010). Teaching children with autism to play a video game using activity schedules and game-embedded simultaneous video modeling. Education & Treatment of Children, 33, 351–370.

Charlop-Christy

M. H.

Daneshvar

(2003). Using video modeling to teach perspective taking to children with autism. Journal of Positive Behavior Interventions, 5, 12–21.

Charlop-Christy

M. H.

Loc

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, 537–552.

Delano

(2007). Video modeling interventions for individuals with autism. Remedial and Special Education, 28, 33–42.

Goldsmith

T. R.

LeBlanc

L. A.

(2004). Use of technology in interventions for children with autism. Journal of Early and Intensive Behavior Intervention, 1, 166–178.

Horner

R. H.

Sprague

Wilcox

(1982). General case programming for community activity. In Wilcox

Bellamy

(Eds.), Design of high school programs for severely handicapped students (pp. 61–98). Baltimore, MD: Paul H. Brookes.

10.

LeBlanc

L. A.

(2010). Using video-based interventions with individuals with autism spectrum disorders: Introduction to the special issue. Education & Treatment of Children, 33, 333–337.

11.

LeBlanc

L. A.

Coates

A. M.

Daneshvar

Charlop-Christy

M. H.

Morris

Lancaster

B. M.

(2003). Using video modeling and reinforcement to teach perspective-taking skills to children with autism. Applied Behavior Analysis, 36, 253–257.

12.

Mechling

L. C.

Ayres

K. M.

Foster

A. L.

Bryant

K. J.

(2013). Comparing the effectiveness of commercially available and custom made videos for teaching cooking skills to high school age students with autism spectrum disorders. Remedial and Special Education, 34, 371–383.

13.

Mechling

L. C.

Gast

D. L.

(2003). Multimedia instruction to teach grocery store location: A study of generalization. Education and Training in Developmental Disabilities, 38, 62–76.

14.

Mechling

L. C.

Gast

D. L.

Krupa

(2007). Impact of SMART Board technology: An investigation of sight word reading and observational learning. Journal of Autism and Developmental Disorders, 37, 1869–1882.

15.

Mechling

L. C.

Hunnicutt

J. R.

(2011). Computer-based video self-modeling to teach receptive understanding of prepositions by students with intellectual disabilities. Education and Training in Autism and Developmental Disabilities, 46, 369–385.

16.

Palechka

MacDonald

(2010). A comparison of the acquisition of play skills using instructor-created video models and commercially available videos. Education & Treatment of Children, 33, 457–474.

17.

Ramdoss

Lang

Fragale

Britt

O’Reilly

Sigafoos

. . . Lancioni

G. E.

(2012). Use of computer-based interventions to promote daily living skills in individuals with intellectual disabilities: A systematic review. Journal of Developmental and Physical Disabilities, 24, 197–215.

18.

Rosenberg

N. E.

Schwartz

I. S.

Davis

C. A.

(2010). Evaluating the utility of commercial videotapes for teaching hand washing to children with autism. Education & Treatment of children, 33, 443–455.

19.

Sancho

Sidener

T. M.

Reeve

S. A.

Sidener

D. W.

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

20.

Sherer

Pierce

K. L.

Paredes

Kisacky

K. L.

Ingersoll

Schreibman

(2001). Enhancing conversation skills in children with autism via video technology: Which is better, “self” or “other” as a model. Behavior Modification, 25, 140–158.

21.

Sindelar

P. T.

Rosenberg

M. S.

Wilson

R. J.

(1985). An adapted alternating treatments design for instructional research. Education & Treatment of Children, 8, 67–76.

22.

Stokes

T. F.

Baer

D. M.

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

23.

Taber-Doughty

Patton

S. E.

Brennan

(2008). Simultaneous and delayed video modeling: An examination of system effectiveness and student preferences. Journal of Special Education Technology, 23, 1–17.

24.

Westling

D. L.

Fox

(2009). Teaching students with severe disabilities (4th ed.). Upper Saddle River, NJ: Merrill Prentice Hall.

25.

Wolery

Gast

D. L.

Hammond

(2010). Comparative intervention design. In Gast

D. L.

(Ed.), Single subject research methodology in behavioral sciences (pp. 329–381). New York, NY: Routledge.