See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/234760906

Using Video Modeling to Increase Variation in the Conversation of Children with

Autism

Article  in  Journal of Special Education Technology · January 2009

DOI: 10.1177/016264340802300305

CITATIONS READS
31 1,453

3 authors, including:

Marjorie H. Charlop Gina Chang

Claremont McKenna College The University of Sydney
86 PUBLICATIONS   3,899 CITATIONS    1 PUBLICATION   31 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Marjorie H. Charlop on 24 February 2015.

The user has requested enhancement of the downloaded file.

 Journal of Special Education Technology

Table of Contents
Topical Issue on Video Technology as an Effective Instructional Tool

I Video Supports for Teaching Students with Developmental Disabilities

and Autism: Twenty- Five Years of Research and Development . . . . . . . . . . . . . . . . . 1
Kevin M. Ayres and John Langone

I Does the Model Matter? Comparing Video Self-Modeling and

Video Adult Modeling for Task Acquisition and Maintenance
by Adolescents with Autism Spectrum Disorders. . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
David F. Cihak and Leigh Schrader

I The Efficacy of Technology Use by People with Intellectual Disability:

A Single-Subject Design Meta-Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Michael L. Wehmeyer, Susan B. Palmer, Sean J. Smith,
Daniel K. Davies, and Steven Stock

I Comparison of Static Picture and Video Prompting on the Performance

of Cooking-Related Tasks by Students with Autism. . . . . . . . . . . . . . . . . . . . . . . . . 31
Linda C. Mechling and Melissa R.Gustafson

I Using Video Modeling to Increase Variation in the Conversation . . . . . . . . . . . . . 47

of Children with Autism
Marjorie H. Charlop, Laura Gilmore, and Gina T. Chang

JSET 2008 Volume 23, Number 3 i

 Journal of Special Education Technology

Editorial Policy and Goals

The Journal of Special Education Technology (JSET) is a to assist individuals with exceptional educational
refereed professional journal that presents up-to-date needs.
information and opinions about issues, research, policy, • Encouraging the development of new
and practice related to the use of technology in the field applications, technologies, and media that can
of special education. JSET supports the publication of benefit individuals with exceptionalities.
research and development activities, provides technologi- • Disseminating relevant and timely information
cal information and resources, and presents information through professional meetings, training programs,
and discussion concerning important issues in the field of and publications.
special education technology to scholars, teacher educa- • Coordinating the activities of educational and
tors, and practitioners. The mission of JSET is: “to provide governmental agencies, business, and industry.
a vehicle for the proliferation of information, research, and • Developing and advancing appropriate technical
reports of innovative practices regarding the application of standards.
educational technology toward the development and educa- • Providing technical assistance, in-service, and pre-
tion of exceptional children.” JSET is a publication of the service education on the uses of technology.
Technology and Media (TAM) Division of the Council • Monitoring and disseminating relevant research.
for Exceptional Children. • Advocating for funds and policies that support the
availability and effective use of technology in this
The goals of TAM include: field.
• Promoting collaboration among educators and • Supporting the activities, policies, and procedures
others interested in using technology and media of CEC and other CEC divisions.

TAM Board Members

Tara Jeffs, President Cheryl Wissick, Treasurer John Castellani, Publications Chair
East Carolina University University of South Carolina Johns Hopkins University
Betty Nelson, President Elect Joel Mittler, CAN Coordinator John Lowdermilk, Professional
University of Alabama–Birmingham Long Island University Development Chair
University of Texas–Pan Am
Deborah Newton, Vice President J. Emmett Gardner, Awards Chair
Southern Connecticut State University University of Oklahoma J. Emmett Gardner, Journal Editor
University of Oklahoma
Brenda Heiman, Past President Diane Painter, Member-at-Large Susan Mistrett, Webmaster
Louisiana Tech University Hood University SUNY at Buffalo
Donna Tilley, Secretary Elizabeth Lahm, Member-at-Large Joy Zabala, Newsletter Editor
PCI Educational Publishing CESA#6, Wisconsin Assistive Technology and Leadership

Subscriptions and Membership in TAM

The Journal of Special Education Technology (JSET) is a traditional Membership inquiries should be directed to the address below or
print-on-paper publication (4 issues per year) that is sent to sub- a membership brochure may be found online at the TAM Web
scribers and members of the Technology and Media Division of site at: http://www.tamcec.org.
the Council for Exceptional Children (TAM). Subscriptions to ]
SET are available without membership in TAM at the following The Council for Exceptional Children
rates (effective September 1, 2008): 1110 North Glebe Road
Individual domestic mail: $69 per year Suite 300
Institutional: $149 per year Arlington, VA, 22201-5704
Electronic Read Only: $129 per year (email required) 703-620-3663 or 1-888-CEC-SPED toll free
International (Canada): $189 per year TTY: 703-264-9446
International (non-Canada): $209 per year fax: 703-264-9494
http://www.cec.sped.org
All inquiries concerning subscriptions should be sent to:
Exceptional Innovations, Inc.
ATT: JSET, P.O. Box 3853, Reston, VA 20195 Copyright © 2008, Technology and Media Division (TAM) of
703-709-0136 the Council for Exceptional Children
[email protected]

ii JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Video Supports for Teaching Students with Developmental

Disabilities and Autism: Twenty- Five Years of Research and
Development
Kevin M. Ayres
John Langone
The University of Georgia
Guest Editors

Educators have used video technology in a variety of ways to address the educational needs of students with
and without disabilities. This article addresses some of the historical and theoretical perspectives related to
video to broadly illustrate how the use of video has improved educational opportunities for students with dis-
abilities. Successful instructional and assistive technology applications of video are discussed, and ideas about
the future of video are offered. The article closes with an exploration of the impact of market and financial
forces that pose challenges for teachers eager to access high-quality video based instructional materials, and
how most applications involve teachers or researchers taking the initiative to create the materials on their
own. The experimental studies and scholarly articles included in this topical issue of the Journal of Special
Education Technology demonstrate that the instruction benefits of video-based technology hold tremendous
potential for students with more significant disabilities.

T eachers have long understood the potential effec-

tiveness of using some type of visual and/or au-
dio support to accompany their verbal presentations to
using the linear technology afforded by the VHS tape
player. Laser disc technology not only provided outstand-
ing video and sound, it also allowed for a sophisticated
students. From attempts at using photographic slides, search and retrieval system. This capability was enhanced
filmstrips, reel-to-reel films, and other examples of tech- when Hypercard software on early Macintosh computers
nology support, teachers have continuously worked to was able to interface with laser disc players and access
provide their students with assistance in mastering dif- portions of the video virtually instantaneously, thus al-
ficult concepts associated with academic content from lowing for computer-based presentations to be linked to
science to English literature. As technology has become supporting video clips.
more advanced over the years, the delivery of the visual/
auditory supports has become more efficient. From reel- This topical edition of the Journal of Special Education
to-reel videos shown in many classrooms to help students Technology presents a sampling of articles by researchers
learn a variety of concepts, a movement to improve video who are engaged in cutting-edge work that illuminates
supports hit a high point when laser disc players became ways to make video technology more effective as an in-
a popular fixture in many general education classrooms structional tool. We attempted to choose authors who
in the 1980s and ’90s. are working on unique applications of video technology
while also providing an article that looks at the bigger
Laser disc players offered a tremendous advantage for picture of how technology as a whole is being used and
teachers because the technology allowed for virtually in- evaluated by the field. For our contribution to this edi-
stant access to any part of the video. Up to this point, tion, we are providing a brief summary of some of the
teachers who wished to return to a spot in a video to em- highlights related to the use of video for helping persons
phasize an important concept or enhance it with further with disabilities become more independent as well as a
discussion had to fast forward or returned to the spot summary of some of the information we have gained

JSET 2008 Volume 23, Number 3 1

 Journal of Special Education Technology

over the years as we have engaged in research and devel- a support for teaching mathematics and science (CTGV,
opment at the University of Georgia. 1993, 1994).

The foundation for the development and research con-

A Theoretical and Historical ducted by CTGV was first delineated by Whitehead
(1929) and later described by Brown, Collins, and Dug-
Perspective on Video Supports uid (1989) in a seminal article that changed the way many
The following discussion of some of the theoretical foun- researchers and educators looked at curriculum develop-
dations highlights how early work in situated cognition ment for all students. The theory of situated cognition
helped to legitimize the use of video in the classroom provided a base for understanding how learners can best
and illustrates how video-based instruction (for any tar- acquire abstract and complex knowledge (Brown et al.,
get skill) involves far more than showing a 90-minute 1989). Simply, proponents of situated cognition stressed
documentary DVD or allowing students to watch televi- the need to tie academic concepts to situations that occur
sion in “freetime.” Whether teaching academic skills like in everyday life. For example, teachers of mathematics
math, community skills like shopping, or social skills like who can demonstrate how complex formulas can be used
how to carry on a conversation, the targeted use of video in the building trades stand a better chance of helping
has grown and has empirical support across theoretical their students understand those concepts and generalize
and pedagogical perspectives. those skills to a variety of situations.

The use of video for teaching students with intellectual The researchers at CTGV (1993) adopted the principles
disabilities and autism is founded on a blend of theories outlined in relation to the theory of situated cognition
and philosophical approaches. Many of the social learn- and developed a number of computer-based products
ing principles explicated by Bandura (cf. Bandura, 1977) that incorporated video as a means of assisting students
drive the work of those focused on the potency of video in gaining academic skills. The development of the Jas-
for modeling academic, functional, and social behaviors. per series of educational products shined light on the ap-
Researchers in this area capitalize on the ability to fo- proach that CTGV called anchored instruction (1990).
cus instruction on salient features of a model depicted Anchored instruction is an educational approach that
on a monitor or TV screen. Running parallel is the use uses “video anchors” to set up problem scenarios for stu-
of video from the perspective of situation cognition and dents to work to solve. For example, one of the more
anchored instruction. Work in this area helped to bring famous Jasper episodes entitled Rescue at Boone’s Meadow
video into the classroom as an important instructional provided a short movie about a protagonist by the name
tool. of Jasper. Jasper and his friends embark upon an adven-
ture to save a wounded the eagle. Through the course
The following historical overview follows these two par- of the adventure, the characters interact in a variety of
allel paths and shows how they addressed learner needs events, including flying an ultra-light airplane. The video
though the use of video. In the limited space we have to anchor includes a variety of information related to math
describe this history it is impossible to include all key and science concepts (e.g., computing fuel economy and
empirical studies, but we hope to illustrate the progres- the flying weight of the ultra light vehicle). Students
sion on multiple levels and show how it ultimately ben- working in groups participate in classroom-based prob-
efits students with disabilities. lem-solving activities related to the video anchor.

Anchored instruction. As the laser disc player became Research conducted by CTGV provided evidence that
more popular, researchers began to look for ways to students not only mastered the targeted skills while using
use this technology more effectively and efficiently. The an anchored instructional approach, but also found the
Cognition and Technology Group at Vanderbilt (CTGV, activities motivating (e.g., Vye, Goldman, Voss, Hmelo,
1990) provided the field with the most intriguing re- Williams, & CTGV, 1997). Reports from teachers indi-
search base on the use of this technology. Over a period cated that this type of anchored instruction motivated
of several years, the CTGV produced an impressive body students to seek answers to their questions and assisted
of literature demonstrating the effectiveness of video as

2 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

students in generalizing the skills to other tasks and situ- cial educators and researchers began to develop and test
ations (CTGV, 1992). different video-based instructional products designed to
improve the learning of students with disabilities (Brown-
Over the course of several years CTGV produced and ing, Nave, White, & Barkin,1985; Crusco, Carter, &
tested several of these technology-based products and McGrath,1986). For example, Browning et al. developed
sold them commercially through an outside vendor. a number of computer applications designed to teach life
Unfortunately, as CD ROM and DVD technology have enhancement skills to students with cognitive and physi-
replaced laser disc technology, the Jasper series does not cal disabilities. Similar attempts at program development
appear to have been transferred for use on the personal that incorporated videos and instructional programs be-
computer. gan to appear across the special education literature base
(e.g., Crusco et al., 1986)
Translating and blending some of the ideas that CTGV
advanced with operant learning principles, we have seen Many research studies have used computer-based video
the adoption of video-based technologies for a wide instruction to teach functional skills to persons with
range of students needs. However, effective technologi- moderate and severe intellectual disabilities, including
cally based instructional devices are not widely available those addressing the issue of promoting generalization
because the commercial market does not see a sufficient through the use of simulated video instruction. Cuvo
profit in continuing production, especially for smaller and Katt’s (1992) seminal study focused on reading in-
populations with highly specific needs. struction for persons with disabilities through the use of
video recordings. Their research compared the effects of
This is particularly apparent in the studies we present community-based instruction, video-based instruction,
in this issue. Much of the research includes fewer par- and flashcard presentation on the acquisition of func-
ticipants, and because researchers are working with rela- tional sight words. Results showed acquisition and gen-
tively small populations of students (compared to entire eralization of words across all three methods. This effort
schools of children working on general education curric- is one of the few that investigated the relationship be-
ulum), the dilemma of producing technology for smaller tween traditional classroom-based activities, instruction
markets becomes more apparent. When this happens, it in the community, and a technology-based simulation.
becomes the responsibility of the teacher to create and The findings were promising and provided evidence that
generate high-quality anchored content. This is a theme the video-based instruction can have generalized effects.
that we will return to later in this article, since it has
significant ramifications for technology applications for Several studies have attempted to assess the effectiveness
individuals with more severe cognitive disabilities. of video or photographs to support the instruction of
community-referenced behaviors to individuals with
Multimedia supports and functional skills instruction. moderate to severe intellectual disabilities and autism.
Some of the earliest development and research in the use Wissick was one of the first to see the potential of using
of video-based technology for instruction was targeted laser disc-based video controlled by computer-assisted
for individuals who have significant hearing problems, instruction as a support for community-based activities
including Dillingham, Roe, and Roe (1982) and Pol- (i.e., grocery shopping). She used an interactive video-
lard (1985). Dillingham et al. used videotape presenta- based computer simulation program to teach three ado-
tions (a new technology at the time) in conjunction with lescents with moderate intellectual disabilities to move
computer-based instruction for teaching students with down aisles and select target items on the computer
various hearing impairments a variety of educational screen (Wissick, Lloyd, & Kinzie, 1992).
content. Pollard used videotape presentations to support
computer-based instruction designed to teach receptive As a basis for their line of research, Haring, Kennedy, Ad-
sign language skills. ams, and Pitts-Conway (1987) used videotapes to teach
three students with autism to locate an item in one store
During the 1980s, a number of creative developments and then measured generalization to other stores in the
took place relating to interactive video instruction across community. One interesting component of the Haring
various educational fields (McNeil & Nelson 1991). Spe- study was that video captions were used to assist students

JSET 2008 Volume 23, Number 3 3

 Journal of Special Education Technology

in generalizing skills (e.g., location of items). Alcantara ing manipulative supports, or items representing real life
(1994) also used video captions to teach location of one situations (e.g., cereal boxes). Simulations that use real
item at a time to three children with autism. Results of materials (e.g., empty boxes) in non-realistic environ-
this study indicated that combining instruction in the ments (e.g., classrooms) may serve to fill in schedules but
community and video training produced improved levels may not improve a student’s ability to shop in a grocery
of performance across stores. store. This problem may be due to the fact that traditional
instructional methods (e.g., flashcards used in classroom-
Based on a number of research efforts by different groups, based problem solving activities) do not typically provide
Wissick, Gardner, and Langone (1999) suggested that learners with a sufficient number of exemplars depicting
the more similar the simulation (video or computer- or referenced to actual community contexts. In short, we
based) is to the target setting, the greater the likelihood need to explore computer-based and video-supported
that transfer will occur. These findings across decades of strategies that supplement actual instruction in realistic
applied research support the use of common stimuli as a environments where the actual skills occur. The popular
strategy for promoting the generalization of skills from press reveals that the use of driver trainer simulators or
training to community settings. Generalization is more flight simulators can cut down on the amount of time
likely when the common stimuli are presented within spent in actual field exercises while improving decision
numerous learning trials (Stokes & Baer 1977, Stokes & making and psychomotor skills. However, reports indi-
Osnes, 1989). Therefore, in terms of video modeling and cate that the simulations cannot completely replace the
computer-based simulation of functional skill instruc- real thing.
tion, Langone, Clees, Reiber, and Matzko (2003) rec-
ommend (a) using common physical stimuli (real cans The computer can be a powerful tool for facilitating gen-
or packages, or photographic/video representations of eralization of skills via instruction anchored to multime-
them, rather than unrealistic drawings on worksheets); dia cues. The processing capabilities of today’s desktop
(b) using common social stimuli (replicating verbiage computers afford many possibilities for the design of
used by store employees); (c) using many examples of var- highly visual and highly interactive learning environ-
ied stimuli that precede the same response class (teaching ments, and, as related to generalization, can result in a
a student to select numerous grocery items); (d) using significant increase in the number of learning trials avail-
many examples of correct responses to the same stimuli able. Such computer-based learning environments are
(teaching a student numerous greetings); and (e) using typically referred to as interactive multimedia (Alessi &
natural contingencies (e.g., selecting the correct change Trollip, 2001).
to buy a can of soda results in the buying of the soda).
Our work has been designed to extend this line of re-
search by investigating whether video-based models, in
Common Sense Approach to the Use conjunction with computer-based activities, can yield
similar results. Some of the our work and the work of
of Video for Supporting Instruction our graduates (e.g., Ayres & Langone, 2002; Ayres &
One of the most important conclusions we have wrestled Langone, 2007; Ayres, Langone, Boon, & Norman,
with is that we cannot become enamored with technol- 2006; Langone, Shade, Clees, & Day 1999; Mechling
ogy (at least for the foreseeable future) as the end-all for & Langone, 2000, Mechling, Gast, & Langone, 2002;
improved instruction of students with moderate to severe Simpson, Langone, & Ayres, 2004) indicates that picto-
cognitive disabilities and autism. Anchored instruction rial and video models as well as interactive video hold
has proven to be a successful way of providing context, promise for teaching individuals with moderate to se-
organization, and problem solving skills. Unfortunately, vere disabilities community-based functional and social
there is still little evidence that the skills learned via visual skills. This work and the work of others demonstrates
anchors used in isolation (e.g., picture cards) generalize to that pictorial and video models assist learners in acquir-
actual tasks in the community. The lack of generalization ing target skills and in generalizing those target skills to
may be exacerbated if the term anchored is used in a more community settings. However, these efforts are limited
general way to mean linking instruction of concepts and due to prohibitively high costs associated with develop-
skills to traditional classroom-based materials such as us- ing video scenes of sufficient quality to provide learners

4 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

with realistic video portrayals of complex functional and system also functions as a cordless pointer that can be
social skills. used to select items on the screen. Physical rehabilitation
centers have begun using this gaming system to keep pa-
tients more interested in the routines of their therapy. A
The Future of Video Supported system like the Wii potentially could allow students to
point, pick up, move, place, and rearrange things much
Instruction like they might do at a job site. Via a computerized sys-
What lies beyond video? What will the next 10 years bring tem, an unlimited range of exemplars and environments
in terms of blending technology and video as an educa- could be presented that might not be logistically possible
tional medium for students with developmental disabili- to visit in a student’s community.
ties? Researchers and teachers have been using television
monitors and computers (desktop and laptop) to provide Unfortunately, these two examples of what “could be” are
students with video models and video-based instruction very unlikely to happen, since the cost and programming
for decades. The future will undoubtedly include the use skill required to execute such projects would unlikely see
of video on portable DVD players, SmartBoards (e.g. a positive financial return. At the American Association
Mechling, Gast, & Krupa, 2007), and other video de- of Mental Retardation and Developmental Disabilities
livery media (perhaps SmartPhones). As we progress, in- conference in 2003, David Braddock addressed this mar-
vestigators will be able to use those features of video and ket-driven issue concerning technology development for
modeling that are most important for learning, and this individuals with severe cognitive disabilities. As much as
will inform practice as teachers try to identify whether those of us in the field see a need for these types of prod-
the best model is one that is video based or static, if it ucts, Braddock provided the simple reality check that for
might vary by task (e.g., Cihak, Alberto, Taber-Doughty, software companies there is not a sufficiently large mar-
& Gama, 2006; Mechling & Gustafoson, 2008), or if ket for these products to justify the cost of development.
perhaps video self-modeling is more powerful than tra- Even though there are people who will benefit from the
ditional video modeling formats (Cihak & Schrader, wonderful things many professionals envision, corpora-
2008). tions have an obligation to make profits, and there are
not enough individuals with severe cognitive disabilities
To forecast the next step with video models and simula- to make it worth their efforts.
tion would be to place a fool’s wager on the next turn
that technology will take. While it is truly impossible Braddock, in the same presentation, posited that we need
to predict technology innovation beyond a few years, to find ways to combine the technology requirements of
indulging in some speculation allows us to begin to set persons with disabilities with the technology needs of an-
new goals and objectives. Certainly, there are more op- other group of individuals, such as the elderly. Braddock
portunities in (video) games for enhancing instruction. believes that the increase in market options by combin-
One advance that is sure to become a reality is the use of ing the needs of these two groups will be more attractive
first-person simulators whereby learners manipulate the for developers and production companies to produce
actions as if they were the main characters in a computer and market promising assistive technology solutions.
game. A popular genre of commercial video gaming al-
ready uses this model to allow users to interact with vir- Braddock’s idea is an excellent one and it should be
tual environments. pursued as researchers and developers enter into more
collaborative relationships. Combining different groups
In a similar vein, the development of motion sensing of people who might jointly benefit from assistive tech-
interactive games may hold promise for allowing stu- nology makes sense when considering advancements in
dents to engage in behavioral rehearsals in a controlled tools and devices such as environmental controls, sight
virtual environment. The Nintendo Wii, which came to enhancements, home and community safety related de-
the U.S. market in 2006, allows users who hold motion- vices, and navigation/mobility tools. Unfortunately, it
sensing controllers to interact with the game by simulat- becomes more difficult to find compatible groups that
ing movements such as swinging a baseball bat, throwing might benefit from software for classroom use that can
a football, rolling a bowling ball, and driving a car. The assist learners with more significant disabilities to attain

JSET 2008 Volume 23, Number 3 5

 Journal of Special Education Technology

and generalize skills related to independent living, func- but also to allow teachers and caregivers across the coun-
tional communication, and social skills. The type of soft- try to share what they have created for their students. File
ware and delivery systems that will provide the best video sharing is not a unique step forward for the use of video
and animated setting will be costly in terms of develop- for instructional purposes. Rather, it is the 21st century
ment. This may be a reason why the quality of video and version of sharing lesson plans or worksheets. User cre-
other multimedia supported software designed for these ated and controlled content is found in many other on-
learners is poor. Being a low-incidence group, school- line venues.
aged learners with significant disabilities simply do not
provide a large enough market, as opposed, for example, Another possibility that holds promise for students,
to a group of low-achieving learners with reading chal- teachers, and teacher educators is Second Life by Linden
lenges. Companies are much more willing to provide de- Labs. We do not have sufficient space to fully explain
velopment funds for reading programs with high-quality Second Life and its potential for instruction. However,
instructional design and multimedia supports that can it is quite literally, a second world. Within the world of
take advantage of potentially large numbers of users. Second Life, users create avatars and then move around
This is not an indictment, but merely the reality of a free and interact with other (real) people. Second Life poten-
market society. What this means for professionals is that tially has all of the good and the bad of this world, but it
we will have to increasingly create high-quality materials may provide a unique opportunity.
with our own video cameras and computers that will ad-
dress the needs of students with disabilities and then find Linden Labs has been working with universities to create
ways to share these materials via the Internet. virtual classrooms where faculty and students can interact
in a very visual environment. Imagine for a moment cre-
As evidenced by the articles in this issue, all of the video ating virtual street crossing to get students used to look-
“products” are tools and instructional strategies that the ing both ways in traffic and to practice estimating the
researchers themselves developed. Available software ti- speed of cars as they approach. Imagine creating a virtual
tles that incorporate video and realistic images to provide office where students with severe cognitive disabilities
video-based or video-simulated instruction are miniscule could interact socially with others in their environments
compared to the wealth of options to teach children to and witness the virtual/real outcomes of appropriate and
read. For this reason, one has to wonder if the next evo- inappropriate social actions. In Second Life, develop-
lution of video technology and simulation will not come ers are only limited by their imagination and the time
from within the field. Innovators capitalizing on existing they want to invest in “building” new environments. For
systems as they have in the past (e.g., HyperCard, Hy- example, John Lester, a researcher at Harvard Medical
perStudio, Authorware, Flash, PowerPoint), stand ready School “founded” Brigadoon as a place for individuals
to develop the next “new” tool that can help individuals with Asperger’s Syndrome to interact online. The setting
with significant disabilities to more effectively reach their provides a controlled environment to interact and create
potential. social opportunities.

This discussion is directly linked to contemporary buzz- What might happen if educators coalesced to construct
words like “user-generated” and “social networking.” virtual learning environments to benefit students with
For example, teachers could share their materials via the moderate to severe developmental delays and autism?
searchable database provided by YouTube. By “tagging” Certainly, we would have to concern ourselves with gen-
videos and relying on user feedback to identify quality eralization, but we have to do that with our current tech-
videos, teachers could quickly tap into a wealth of video nologies and we have been documenting generalization
supports for their students. Similarly, teachers with com- from computer-based instruction to real life environ-
mon needs may become “friends” via social network tools ments already (see Ayres, Langone, Boon, & Norman,
such as LinkedIn, FaceBook, or MySpace and through 2006). One of the appealing things about a setting like
these avenues connect with others who generate content. Second Life or YouTube is that it is cooperative. That is,
These group-supported venues have the potential not we can pool our talents for the greater good. Those who
only to advance the use of video as an instructional tool, are motivated by profits might find a place in Second

6 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Life, but development does not have to be market driv- References

en. Researchers and teachers have never waited for the
Alcantara, P. R. (1994). Effects of videotape instructional package on
market to drive the development of instructional prod- purchasing skills of children with autism. Exceptional Children,
ucts, and for the group of individuals with moderate to 61, 40-55.
severe developmental disabilities, we do not want to miss Alessi, S., & Trollip, S. (2001). Multimedia for learning: Methods and
an opportunity to develop this and other potential strate- development (3rd ed.). New York: Allyn & Bacon.
gies and devices. This idea is not new, but it represents Ayres, K. A. & Langone, J. (2002). Acquisition and generalization of
what teachers of students with significant developmental purchasing skills using a video enhanced computer-based instruc-
disabilities have been doing for a long time (i.e., creating tional program. Journal of Special Education Technology, 17(4),
their own materials and sharing them with their friends 15-28.
and colleagues who have similar needs). Now we have Ayres, K. M., & Langone, J. (2005). Using video for instruction of
technologies that allow us to expand our reach, enlarge students with autism: A review of the literature. Education and
Training in Developmental Disabilities, 40, 183-196.
our library, and share our collective knowledge, skills, and
time to generate video- and technology-based materials Ayres, K., & Langone, J. (2007). A comparison of video modeling
perspective for students with autism. Journal of Special Education
that can support and augment traditional instruction.
Technology, 22(2), 15-30.

You will see from the experimental studies and scholarly Ayres, K. M., Langone, J., Boone, R., & Norman, A. (2006). Com-
puter based instruction for purchasing skills. Education and Train-
articles included in this topical issue of JSET that video- ing in Developmental Disabilities, 41, 253-263.
based technology holds tremendous potential benefits for
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ;
students with more significant disabilities. Whether it is Prentice-Hall.
interacting with video models of social scripts (Charlop-
Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition
Christy, Gilmore, & Chang, 2008) to enhance commu- and the culture of learning. Educational Researcher, 18, 32-41.
nication skills, comparing video to static picture prompts
Browning, P., Nave, G., White, W. A., & Barkin, P. Z. (1985). In-
to teach cooking (Mechling & Gustafson, 2008), or teractive video as an instructional technology for handicapped
comparing video modeling to video self-modeling to learners: A development and research program. Australia & New
teach prevocational skills (Chiak & Schrader, 2008), Zealand Journal of Developmental Disabilities, 11(3), 123-128.
video materials created by nonprofessionals (i.e., those Charlop-Christy, M. H., Gilmore, L., & Chang, G. T. (2008). Using
without professional training in video production) can video modeling to increase variation in the conversation of chil-
benefit students with developmental disabilities. This is- dren with autism. Journal of Special Education Technology, 23(3),
47-66.
sue also includes an article by Wehmeyer, Palmer, Smith,
Davies, and Stock (2008), who are conducting a federal Cihak, D. F., Alberto, P. A., Taber-Doughty, T., & Gama, R. I. (2006).
A comparison of static picture prompting and video prompting
grant designed to catalog and analyze the literature base simulation strategies using group instructional procedures. Focus
on the use of technology by individuals with intellec- on Autism and Other Developmental Disabilities, 21(2), 89-99.
tual disabilities. Their work includes the development
Cihak, D. F., & Schrader, L. (2008). Does the model matter? Com-
of a database of scholarly and opinion pieces as well as paring video self-modeling and video adult modeling for task ac-
a meta-analysis of articles that present databased stud- quisition of adolescents with autism spectrum disorders. Journal
ies. We think the readers of this issue will discover new of Special Education Technology, 23(3), 9-20.
trends in the research from reading their analysis as well Cognition and Technology Group at Vanderbilt. (1990). Anchored
as get a clearer picture of the gaps in the research that instruction and its relationship to situated cognition. Educational
we should be addressing. As technology evolves, we fully Researcher, 19(6), 2-10.
expect user/teacher-generated content to positively influ- Cognition and Technology Group at Vanderbilt. (1992). The Jasper
ence classroom instruction and by working cooperatively experiment: An exploration of issues in learning and instruction-
al design. Education Technology Research & Development, 40(1),
through the Internet, individual efforts for individual 65-80.
students can have a much broader impact.
Cognition and Technology Group at Vanderbilt. (1993). Anchored
instruction and situated cognition revisited. Educational Technol-
ogy, 33(3), 52-70.
Cognition and Technology Group at Vanderbilt. (1994). The rela-
tionship between situated cognition and anchored instruction: A
response to Tripp. Educational Technology, 34(8), 28-32.

JSET 2008 Volume 23, Number 3 7

 Journal of Special Education Technology

Crusco, A. H., Carter, P., & McGrath, M. (1986). Skill requirements Pollard, G. (1985). The nuts and bolts of interactive video: How it
for interactive video instruction of persons with mental retarda- works and how to get started. American Annals of the Deaf, 130,
tion. Mental Retardation, 24, 99-105. 386-390.
Cuvo, A. J., & Katt, K. P. (1992). Effects of community-based, vid- Simpson, A., Langone, J., & Ayres, K.M. (2004). Embedded video
eotape, and flash card instruction of community-referenced sight and computer-based instruction to improve social skills for stu-
words on students with mental retardation. Journal of Applied Be- dents with autism. Education and Training in Developmental Dis-
havior Analysis, 25, 499-512. abilities, 39, 240-252.
Dillingham, L. M., Roe, J. J., & Roe, M. D. (1982). Selected ap- Stokes, T. F., & Baer, D. M. (1977). An implicit technology of gener-
plications of computer-assisted video instruction in the education alization. Journal of Applied Behavior Analysis, 10, 349-367.
of hearing-impaired students. American Annals of the Deaf, 127,
Stokes, T. F., & Osnes, P. G. (1989). An operant pursuit of generaliza-
652-658.
tion. Behavior Therapy, 20, 337-355.
Haring, T. G., Kennedy, C. H., Adams, M. J., & Pitts-Conway, V.
Vye, N. J., Goldman, S. R., Voss, J. E., Hmelo, C., Williams, S., &
(1987). Teaching generalization of purchasing skills across com-
Cognition and Technology Group at Vanderbilt. (1997). Com-
munity settings to autistic youth using videotape modeling. Jour-
plex mathematical problem solving by individuals & dyads. Cog-
nal of Applied Behavior Analysis, 20(1), 89-96.
nition and Instruction, 15, 435-484.
Langone, J., Clees, T., Rieber, L., & Matzko, M. (2003). The future
Wehmeyer, M. L., Palmer, S. B., Smith, S. J., Davies, D. K., & Stock,
of instruction using interactive technology for individuals with
S. (2008). The efficacy of technology use by people with intel-
moderate to severe disabilities: Issues relating to research and prac-
lectual disability: A single-subject design meta-analysis. Journal of
tice. Journal of Special Education Technology, 18(1), 5-16.
Special Education Technology, 23(3), 21-30.
Langone, J., Shade, J., Clees, T. C., & Day, T. (1999). Effects of
Whitehead, A. N. (1929). The aims of education. New York:
multimedia instruction on teaching functional discrimination
MacMillan.
skills to students with moderate/severe intellectual disabilities. In-
ternational Journal of Disability, Development and Education, 46, Wissick, C. A., Gardner, J. E., & Langone, J. (1999). Video-based
493-513. simulations: Considerations for teaching students with develop-
mental disabilities, Career Development and Exceptional Individu-
McNeil, B. J., & Nelson, K. R. (1991). Meta-analysis of interactive
als, 22, 233-249.
video instruction: A ten-year review of achievement effects. Jour-
nal of Computer-Based Instruction, 18(1), 1-6. Wissick, C. A., Lloyd, J. W., & Kinzie, M. B. (1992). The effects of
community training using a videodisc-based simulation. Journal
Mechling, L. C., Gast, D., & Krupa, K. (2007). Impact of SMART
of Special Education Technology, 11(4), 207-222.
Board technology: An investigation of sight word reading and
observational learning. Journal of Autism and Developmental Dis-
abilities, 37, 1869-1882. Author Notes
Mechling, L. C., Gast, D., & Langone, J. (2002). Computer-based
video instruction to teach persons with moderate intellectual dis- Dr. Kevin M. Ayres is an assistant professor, Department of Com-
abilities to read grocery aisle signs and locate items. Journal of Spe- munication Sciences & Special Education at The University of
cial Education, 35, 224-240. Georgia. Dr. John Langone is a professor, Department of Com-
munication Sciences & Special Education at The University of
Mechling, L. C., & Gustafson, M. R. (2008). Comparison of static
picture and video prompting on the performance of cooking re- Georgia.
lated tasks by students with autism. Journal of Special Education Correspondence should be addressed to Kevin M. Ayres, Depart-
Technology, 23(3), 31-45.
ment of Communication Sciences & Special Education, The Uni-
Mechling, L. C. & Langone, J. (2000). The effects of a computer- versity of Georgia, 557 Aderhold Hall, Athens, GA 30606. Email
based instructional program with video anchors on the use of to [email protected].
photographs for prompting augmentative communication. Edu-
cation and Training in Mental Retardation and Developmental Dis-
abilities, 35, 90-105.

8 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Does the Model Matter? Comparing Video Self-Modeling and

Video Adult Modeling for Task Acquisition and Maintenance
by Adolescents with Autism Spectrum Disorders
David F. Cihak
Leigh Schrader
University of Tennessee

The purpose of this study was to compare the effectiveness and efficiency of learning and maintaining vo-
cational chain tasks using video self-modeling and video adult modeling instruction. Four adolescents with
autism spectrum disorders were taught vocational and prevocational skills. Although both video modeling
conditions were effective for acquiring and maintaining vocational skills, one participant performed more
effectively during video self-modeling instruction. Two participants acquired skills more efficiently during
video self-modeling instruction and the fourth participant showed no functional preference for self-model-
ing or adult modeling conditions. In addition, the classroom teacher indicated that video modeling methods
were socially acceptable for school settings and would recommend them to other teachers.

T he use of video is a relatively new genre of visually

based treatments that hold promise for individuals
with disabilities. Video modeling consists of showing the
feldt, 2004; Cihak, Alberto, Taber-Doughty, & Gama,
2006; Haring, Breen, Weiner, Kennedy, & Bednersh,
1995), grocery shopping (e.g., Kyhl, Alper, & Sinclair,
participant a video of someone performing a target skill 1999; Mechling, Gast, & Langone, 2002), conversa-
or behavior. After viewing the entire video, the partici- tion skills (e.g., Sherer et al., 2001), and play skills (e.g.,
pant is provided an opportunity to perform the target D’Ateno, Mangiapanello, & Taylor, 2003; Reagon, Hig-
behavior (Alberto, Cihak, & Gama, 2005). If the par- bee, & Endicott, 2006).
ticipant commits an error, the video is shown a second
time (Le Grice & Blampied, 1994; Tiong, Blampied, & Video modeling procedures are rooted in Bandura’s
Le Grice, 1992). Video modeling has been used success- (1997) social learning theory, which states that individ-
fully to teach individuals with intellectual disabilities uals can attain a behavior by watching a “model” (the
meal preparation (e.g., Cannella-Malone et al., 2006; person demonstrating the behavior) rather than through
Graves, Collins, Schuster, & Kleinert, 2005; Rehfeldt, direct personal experience. Models influence learning in
Dahman, Young, Cherry, & Davis, 2003; Sigafoos et al., observers through their ability to convey information.
2005; Van Laarhoven & Van Laarhoven-Myers, 2006), The observer acquires information from a model by
daily living skills (e.g., Branham, Collins, Schuster, & watching and attending to the model. From this sym-
Kleinert, 1999; Norman, Collins, & Schuster, 2001; Van bolic representation, a person can retrieve information
Laarhoven & Van Laarhoven-Myers, 2006), purchasing at any time and use it to guide behavior. However, to ac-
(e.g., Alberto, Cihak, & Gama, 2005; Bidwell & Reh- quire a new skill, individuals must attend to the model.

JSET 2008 Volume 23, Number 3 9

 Journal of Special Education Technology

Individuals with autism are known to favor visual treat- type of model to use is warranted, given that various
ment approaches (e.g., MacDuff, Krantz, & McClanna- characteristics of a model may influence attention. The
han, 1993; Pierce & Schreibman, 1994). Hermelin and most effective models were those that were perceived
O’Connor (1970) reported that children with autism en- by observers to be similar to themselves (e.g., age, gen-
code visuospatial information better than auditory infor- der), as well as competency, likeability, and social sta-
mation. Their ability to encode and organize information tus (Bandura, 1997). Using individuals to act as their
was found to be influenced by the duration for which the own models was proposed as a method to maximize ob-
stimuli remained. Later evidence from intelligence test- server characteristics that serve to promote attention to
ing revealed that children with autism performed better the model (e.g., Brown & Middleton, 1998; Hosford,
on tasks such as form discrimination, matching, block 1981). Video self-modeling also may increase self-effi-
design, object assembly, and pattern analysis, all of which cacy and the reinforcing value of watching oneself being
involved the stimuli remaining visible (e.g., Harris, Han- successful (Bray & Kehle, 1996). Norman et al. (2001)
dleman, & Burton, 1990; Siegel, Minshew, & Goldstein, used video self-modeling to teach students with autism
1996). Moreover, children with autism, with both high functional skills. Other types of models also were used
and low IQ scores, performed better on visuospatial tasks successfully in enhancing task acquisition, including an
than tasks requiring social or language reasoning (Lin- adult without disabilities (e.g., D’Ateno et al., 2003), an
coln, Courchesne, Kilman, Elmasian, & Allen, 1988). adult with disabilities (e.g., Rehfeldt et al., 2003), peers
Visual supports were also used successfully for children without disabilities (e.g., Branham et al., 1999), siblings
with autism to maintain attention, understand spoken (e.g., Reagon, Higbee, & Endicott, 2006), and the first-
language, and sequence and organize their environment person perspective (e.g., Cihak et al., 2006).
(Hodgdon, 1995). When instruction is facilitated with
visual cues, individuals with autism can acquire, main- Several studies have investigated possible differences be-
tain, and generalize a range of academic (e.g., MacDuff tween the types of video models used to teach students
et al., 1993), functional (e.g., Vaughn & Horner, 1995), with ASD discrete behaviors. For example, Ihrig and
domestic/self-care (e.g., Pierce & Schreibman, 1994), Wolchik (1988) examined the effects of using peers or
leisure (e.g., D’Ateno et al., 2003), and communication adults as video models to teach students with ASD to
(e.g., Bondy & Frost, 1994; Krantz & McClannahan, answer a series of discrete questions. Results indicated
1998) skills. that the students learned equally well with either model.
Moreover, high levels of generalization across settings
Bellini and Akullian (2007) recently conducted a me- and maintenance occurred in both conditions. Likewise,
ta-analysis of the video modeling literature on children Sherer et al. (2001) noted functional similarities using
and adolescents with autism spectrum disorders (ASD). peers or self as video models to answer a series of discrete
Twenty-three single-subject designs from 1987-2005 conversational questions. Ayers and Langone (2007) also
were evaluated for effectiveness, with a special focus on examined different types of video viewing perspectives
video modeling and video self-modeling strategies. Both for students with ASD. Students were taught a discrete
types of strategies were found to be effective for teaching skill of putting grocery items away in proper locations
social, communication, and functional skills. Addition- using either first-person or third-person perspective vid-
ally, video modeling successfully reduced self-stimulatory eos. Both perspectives were analogous. When perform-
behaviors (Brown & Middleton, 1998) and problematic ing discrete skills, previous studies have suggested no
behaviors, including tantrums, aggression, and off-task overall differences in skill acquisition among first-person
behavior (Buggey, 2005; Coyle & Cole, 2004). perspective, video self-modeling, video adult models, or
video peer models.
Although video modeling produced efficacious out-
comes, Bellini and Akullian (2007) warned that con- When providing systematic instruction, a practitioner
tinued research is needed before definitive conclusions may focused on specific, measurable responses that may
can be drawn. The instructional quality components and either be discrete (a single behavior) or a response chain
the socially valid uses of video modeling are of special (e.g., a series of behaviors, such as a task analysis). While
interest. Since attention is essential for observational teaching single units of behavior is effective for teach-
learning (Bandura, 1977), greater investigation of the ing specific skills, sequential or chained responses (e.g.,

10 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

self-help, community, vocational skills) make greater opmental Profile-II indicated an IQ equivalence score of
cognitive, motor, and/or communicative demands on 50, suggesting a moderate intellectual disability. Phil was
the individual. For example, in a behavior chain, the an- 17 years old. Phil also scored in the severe autism range
tecedent stimulus and the response operate so that each on the GARS, and the Developmental Profile-II indicated
response produces a discriminative stimulus for the next an IQ equivalence score of 30, suggesting a severe intel-
response, and so forth. Performance of each discrete re- lectual disability. Finally, Alan was 20 years old. Results
sponse only serves as one individual component of the of the GARS placed Alan within the severe autistic range.
entire skill. In their review of the video instruction litera- His Developmental Profile-II indicated an IQ equivalence
ture on individuals with ASD, Ayres and Langone (2005) score of 35, suggesting a severe intellectual disability.
noted, “Systematically isolating video components and
then repackaging them into the most efficient and effec-
tive tools for teaching should be the goal of research in
Setting
this area” (p. 195). Participants viewed videos of either themselves or an
adult performing a task in the natural setting where the
The purpose of the present research study was to rep- task was typically performed. Ryan and Ronald used a
licate the Sherer et al. (2001) research and to compare copy and fax machine located in the teachers’ workroom.
the effectiveness and efficiency of the type of model used The school had one designated workroom for teachers
on the acquisition and maintenance of chain tasks. Spe- that included three copy machines and one fax machine,
cifically, our study examined whether there were func- as well as other equipment, materials, and resources for
tional differences in learning and maintaining chained teachers. Ryan and Ronald used the same copy machine
vocational skills when video self-modeling was compared and fax machine during all phases of the study. Alan and
to adult modeling for adolescents with ASD. The study Phil packaged first aid kits and prepared “family packs”
also explored the use of a response-based prompting er- in the school’s vocational lab.
ror correction procedure and the social validity of video
modeling methods.
Tasks
Two vocational and two prevocational tasks were selected
based on their similarity and because they were consis-
Method tent with participants’ individual transition plans. Table
1 lists the task analysis for vocational tasks.
Participants
Ryan and Ronald were required to independently com-
Four participants diagnosed with ASD and intellectual
plete 11 steps to make copies and 12 steps to send a fax.
disabilities participated in this study. Participants were
Both skills required fine motor movements to operate a
diagnosed by their private physicians and received pa-
keypad. Phil and Alan packaged first aid kits and fam-
rental permission to participate in the study. All students
ily packs. It took 10 steps to complete both tasks. The
met the following criteria: (a) no prior training with of-
faxing/copying and packaging tasks involved similar mo-
fice tasks or packaging activities, (b) no hearing or vi-
tor movements of reaching, grasping, and placing items
sion impairments that impeded video instruction, (c) no
into a bag or container. Additionally, to help match to
physical disability that impeded the performing of tasks,
sample each participant was provided an index card that
and (d) agreement to participate in the study.
contained the fax number or PIN (to operate the copy
machine).
Ryan was 21 years old. Results of the Childhood Autism
Rating Scale (CARS) (Schopler, Reichler, & Renner,
1988) placed Ryan within the severe autistic range. The Materials and Equipment
Developmental Profile-II (Alpern, Boll, & Shearer, 1988)
Packaging Tasks. For the packaging activities, family
indicated an IQ equivalence score of 45, suggesting a
packs and first aid kits, participants were provided all the
moderate intellectual disability. Ronald was 16 years old.
necessary items to complete all task-analyzed steps (see
The Gilliam Autism Rating Scale (GARS) (Gilliam, 1995)
Table 1). Items were stored in separate boxes on a large
placed Ronald within the severe autistic range. The Devel-

JSET 2008 Volume 23, Number 3 11

 Journal of Special Education Technology

Table 1

Task Analysis for Vocational Tasks

Preparing Preparing
Family Packs First Aid Kits Making Copies Sending a Fax
1. Stack four plates 1. Get plastic container 1. Go to the copier 1. Go to the fax machine
2. Place four napkins on 2. Place four band-aids in 2. Open the lid 2. Get the cover sheet
plates bag 3. Place the paper face 3. Write who the fax is to
3. Place four forks on plates 3. Place four alcohol wipes down 4. Write who the fax is from
4. Touch only the handles in bag 4. Align paper to the top 5. Write what the fax is
of the utensils 4. Place four Q-tips in bag left corner about
5. Place four spoons on 5. Place four gauze pads in 5. Close the lid 6. Write the number of
plates bag 6. Press number of copies pages being sent
6. Place four salt packets on 6. Place four plastic gloves 7. Press start 7. Place the fax face down
plates in bag
8. Remove copies 8. Dial #9
7. Place four pepper packets 7. Place one Ace bandage in
on plates bag 9. Open lid 9. Dial the fax number
8. Open the plastic bag 8. Place one pair of tweezers 10. Remove original paper 10. Press fax button
9. Place the family pack in bag 11. Close lid 11. Wait until screen says
inside the bag 9. Close the bag “OK”
10. Set aside 10. Set aside 12. Remove fax

work table. Participants stood at the work table to pack- screen for additional options. Also, 10 buttons were next
age either the family packs or first aid kits. to the keypad: start, stop/pause, clear all, features, job
status, machine status, help, language, access A, and in-
Faxing/Copying Tasks. A Xerox Pro 657 fax machine was terrupt. The start button was color-cued green, the stop/
used to create the videos and instruct participants on pause was red, the clear all was yellow, and the remain-
sending a fax. The machine included 24 buttons: 12 but- ing buttons were gray. The touch screen consisted of four
tons consisting of zero through nine keypad with pound components: basic copy, image adjustment, output for-
and asterisk buttons, and buttons to start, stop, cancel, mat, and job assembly. Each component displayed six
copy, redial, speed dial, manual dial, manual group, reso- additional options such as under basic copy the student
lution, and contrast. The keypad and start buttons were could select output, top tray collate, main tray collate,
color-cued green and the stop button was color-cued red. 1 staple, 1 staple for legal, and more. Therefore, par-
The remaining buttons were white. Participants were re- ticipants had the option of selecting from 27 different
quired to place the fax cover page and document face functions. Additionally, participants were provided with
down on the top feeder of the fax machine. They were a three-page newsletter and a four-digit PIN written on a
provided with a document to fax and a 10-digit phone 3 x 5 in. index card to operate the copy machine.
number written on a 3 x 5 in. index card to operate the
fax machine.
Video Models
A Xerox 480ST copy machine was used to instruct par- Participants viewed a video of themselves (a self-model
ticipants to make copies. The machine had a top feeder condition) or of an unfamiliar adult (an adult model
tray that required the paper to be placed face-up. The condition) performing the targeted tasks. All videos were
machine consisted of a 12-button keypad and a touch created using a Sony 72X Digital Zoom camera and

12 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

iMovie 4.01 software from iMacintosh G4. The videos Design

were filmed showing the model (adult or self ) perform-
ing the task. Zooming-in was utilized to better show the An alternating-treatment design (Barlow & Hersen,
participants the number pads on the copy and fax ma- 1984) was used to examine the differential effects of self-
chines. For each video, a male voice-over stated each step model and adult model conditions on each participants’
of the task analysis. Participants viewed the videos on a skill acquisition and maintenance performance. The uses
15-in. laptop computer, in either the vocational lab or of self-model and adult model conditions were counter-
the teachers’ workroom. balanced to reduce carryover effects across students and
tasks. Additionally, both conditions were presented on
Self-model videos were created by having the participant an alternating schedule to limit order effects by ensuring
imitate the specific steps of the task immediately after each condition preceded and followed the other.
the teacher’s demonstration. The self-model videos were
edited to show a continuous display of the participant Baseline
demonstrating the entire task correctly and independent-
ly. The self-model videos were produced four weeks prior Baseline procedures occurred either in the teachers’ work-
to baseline data collection. The unfamiliar adult model room or the vocational lab. Participants were provided
was a male in his late twenties of the same ethnicity as with all materials needed to complete the targeted tasks
the participants. prior to beginning the session. They were then instructed
to complete the task (i.e., “make copies, send the fax,
The video showing how to make copies was labeled 33 package a family pack or first aid kit”). No additional in-
s, the video for sending a fax was labeled 34 s. The video structional prompts or feedback were provided. If a par-
for packaging first aid kits was labeled 93 s, and 103 s for ticipant was nonresponsive after 20 s, he was asked to try
preparing family packs. Sound was added for each video to complete any or all steps of the task he may know. Any
using either the participant’s voice or that of the unfamil- task-analyzed step performed independently, regardless
iar male. Participants or the adult model described the of the order, was recorded as independent. If a partici-
motor step required to correctly complete the task (e.g., pant continued to be nonresponsive after one minute, he
“stack four plates, place four band-aids in the bag, press was asked if he was finished. The probe was discontinued
start, and dial the fax number”). if the student confirmed completion of the task. If no re-
sponse continued for an additional 10 seconds, the probe
was discontinued.
Data Collection
Data were collected through the use of a data sheet de- Video Modeling Procedures and Conditions
signed to record the controlled presentation of task ana-
lyzed chains. Both tasks were presented to participants Two different video models were used to teach partici-
once daily during morning and afternoon. Event record- pants how to perform the prevocational or vocational
ing procedures were used to record each step as complet- tasks independently. The physical arrangement was the
ed independently or incorrectly. The mean percentage of same as in the baseline condition, except a laptop com-
independent steps of the task analyses and the number of puter was placed on the worktable or next to the fax and
sessions required to reach acquisition criterion (i.e., three copy machine within view of participants. Participants
consecutive data points with 100% independence) were either watched a video of themselves or of an unfamiliar
recorded during both video conditions. Event recording adult performing the specified task. The classroom teach-
procedures were used to record the total number of er- er pointed to the computer screen, said, “[Name] watch
rors made before reaching criterion. An error was defined this,” and started the video. When the video ended, the
as performing an incorrect step and requiring either a teacher said, “[Name], now you [make copies, send the
second viewing of the video clip or teacher assistance. fax, and package a family pack or first aid kit]”.
After participants met criterion, they were probed three
and six weeks later to determine if the initial instruction Participants received verbal praise after completing the
affected performance over time. task-analyzed steps independently. The acquisition crite-
rion was defined as completing independently all task-an-

JSET 2008 Volume 23, Number 3 13

 Journal of Special Education Technology

alyzed steps for three consecutive sessions. If an incorrect ing the adult model condition. Table 2 lists each partici-
step (error) was performed, the teacher said, “[Name], pant’s assigned tasks and corresponding video modeling
let’s watch the video again,” pointed to the computer conditions.
screen, and played the segment of the video showing the
correct demonstration a second time. If incorrect steps
continued, a least-to-most prompt hierarchy was used to
Maintenance
complete the individual steps of the task until the partici- Follow-up probes were collected three and six weeks
pant performed the correct response without assistance. after the acquisition phase for self-modeling and adult
A four-second interval between prompts was implement- modeling conditions. These probes occurred in the set-
ed. The least-to-most prompt hierarchy consisted of the ting where the participants were initially trained, except
following levels across instructional session: (a) verbal this time they did not view the video model prior to ini-
prompt (e.g., “Do you see the [button to press or item to tiating the task. Maintenance probes were collected to
package], (b) gesture (e.g., pointing to the discriminating determine if the initial instruction affected participants’
stimuli), (c) gesture plus verbal explanation (e.g., point- performance over time.
ing to the discriminating stimuli and providing verbal
explanation), (d) modeling plus verbal explanation (e.g.,
teacher performing the step and providing verbal expla-
Interobserver Agreement and Procedural
nation), and (e) physical assistance plus verbal explana-
Reliability
tion (e.g., holding the participant’s wrist, guiding the Interobserver agreement and procedural reliability data
correct response, and providing verbal explanation). were collected for each participant by a trained second
observer (second author) during 33% of the sessions
The video modeling conditions (self and unfamiliar across phases. The trained second observer recorded
adult) were counterbalanced across tasks. Ryan watched data independently and simultaneously on the steps
himself (self-model condition) sending a fax and watched performed independently or incorrectly by the partici-
an adult model (unfamiliar adult model condition) mak- pant using the same prepared data sheet as the teacher.
ing copies. In contrast, for making copies Ronald was Interobserver agreement was calculated by dividing the
his own model, and he watched the adult model send number of agreements by the number of agreements and
a fax. Phil watched himself packaging first aid kits and disagreements and multiplying by 100%. The resulting
an unfamiliar adult model preparing family packs. Con- mean interobserver agreement was 98% for Ryan, 95%
versely, Alan participated in the self-model condition for for Ronald, and 100% for both Phil and Alan.
preparing family packs and packaged first aid kits dur-

Table 2

Participants, Tasks, Modeling Condition, and Duration of Videos

Participants Tasks Conditions Video Duration

Ryan Making copies Adult model 33 seconds
Sending fax Self-model 34 seconds
Ronald Sending fax Adult model 34 seconds
Making copies Self-model 33 seconds
Phil Preparing a first aid kit Adult model 93 seconds
Preparing a family pack Self-model 103 seconds
Alan Preparing a family pack Adult model 103 seconds
Preparing a first aid kit Self-model 93 seconds

14 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Procedural reliability measures examined the classroom Ryan

teacher’s performance of showing the correct video mod-
el condition per task, ensuring all required materials were Ryan did not send a fax or make copies during baseline.
available to perform the task, using verbal praise contin- During the alternating treatments phase, his mean per-
gent upon steps performed independently, replaying the centage of steps completed independently for sending a
video if errors occurred, and using the correct prompt- fax during the self-model condition increased to 62%,
ing hierarchy and response time. The trained second ob- and to 59% for making copies when the adult was the
server recorded data independently and simultaneously model. Ryan also acquired sending faxes more efficiently
on the teacher’s implementation of intervention proce- during the self-modeling condition, requiring 7 instruc-
dures using a prepared intervention protocol data sheet. tional sessions to reach criterion and making 29 errors.
Procedural reliability was assessed for each participant in During the adult modeling condition, Ryan required 11
33% of the sessions across phases. Procedural reliability sessions to reach criterion, making 55 errors learning to
was calculated by dividing the number of agreements by make copies. The two data path conditions fractionated
the number of agreements and disagreements and mul- between sessions 10 and 15, indicating a preference for
tiplying by 100%. The resulting mean procedural reli- the self-model condition. During follow-up at three and
ability was 100% for Ryan, Ronald, and Phil, and 95% six weeks, Ryan maintained 100% of both skills.
for Alan.
Ronald
Social Validity During baseline, Ronald did not send a fax or make cop-
At the conclusion of the study, participants were asked ies. He increased the mean percentage of independent
which video they preferred (self or adult model). In addi- steps for making copies to 81% during the self-model
tion, the classroom teacher was asked to complete the In- condition and to 70% during the adult model condition.
tervention Rating Profile (IRP)-15 (Martens, Witt, Elliot, Ronald also reached acquisition criterion more efficiently
& Darveaux, 1985). The IRP-15 is a 15-item Likert-type during the self-model condition, requiring 6 sessions and
scale that assesses general acceptability of interventions. making 13 errors. During the adult model condition,
The IRP-15 was adapted to focus on the development, Ronald required 8 sessions to reach criterion, making 29
implementation, and participants’ outcomes of each errors learning to send a fax. During follow-up at three
video modeling condition. The Likert scale ranges from and six weeks, Ronald maintained 100% of both skills.
one—strongly disagree to six—strongly agree. Higher
scores indicate better acceptance of interventions, and Phil
ratings above 52.5 are considered to reflect acceptability
by the rater (VonBrock & Elliott, 1987). Phil did not prepare a family pack or a first aid kit during
baseline. During the alternating treatments phase, the
mean percentage of steps he completed independently
during the self-model condition increased to 46% (family
pack), and to 53% (first aid kit) during the adult model
Results condition. While the mean percentage of independence
Figure 1 shows the percentage of steps completed inde- was higher during the latter condition, Phil reached cri-
pendently by each participant. As illustrated, students ac- terion more quickly during the self-model condition (n =
quired and maintained the skills necessary to use a fax and 13) than the adult model condition (n = 15) although 70
a copy machine, package a first aid kit, or prepare fam- errors occurred during both conditions. Phil also main-
ily packs using video modeling. Both self or adult mod- tained both skills at 100% three and six weeks later.
els were effective. However, Ryan’s data demonstrated a
fractionization between the conditions. Another partici-
Alan
pant (Ronald) acquired the targeted skill more efficiently
during the self-model condition, and two participants During baseline, Alan did not package a first aid kit or
(Phil and Alan) demonstrated no functional differences a family pack. The mean percentage of steps completed
between self-modeling and adult modeling conditions. independently increased to 56% during the self-model
Table 3 lists comparative measures to criterion. condition, and to 55% during the adult model condi-

JSET 2008 Volume 23, Number 3 15

 Journal of Special Education Technology

Figure 1

Figure 1. Percentage of steps completed correctly using video self-modeling and video adult modeling

Baseline Alternating Treatments Follow-up

100 Adult-Model
80 Self-Model
60 Series8
40 Series7
20
0 Ryan
Series5
-20
Series6
1 5 10 15 20 25 30 35

100
Adult-Model
80 Self-Model
Percentage of Steps Completed Independently

60
40
20 Series7
Ronald
0
Series8
-20
1 5 10 15 20 25 30 35

100 Adult-Model
80 Self-Model
60
40
20 Series7
Phil Series5
0
Series6
-20
Series8
1 5 10 15 20 25 30 35

100 Adult-Model
80 Self-Model
60
40
20 Series7
Alan Series5
0
Series8
-20
Series6
1 5 10 15 20 25 30 35

Sessions

16 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Table 3

Comparative Measures to Criterion

Video Self-Model Video Adult Model

Mean Mean
Percent of Session to Number of Percent of Session to Number of
Participants Independence Criterion Errors Independence Criterion Errors
Ryan 62† 7† 29† 59 11 55
Ronald 81† 6† 13† 70 8 29
† †
Phil 46 13 70 53 15 70
† †
Alan 56 11 48 55 11 50
†
Greater efficacy or efficiency indicated.

tion. Alan required 11 sessions to reach criterion for both The teacher’s most salient reason for preferring the adult
conditions; he made 48 errors during the self-modeling model over the self-model was due to the time and effort
condition and 50 errors during the adult model con- required for material development. Specifically, using the
dition. During follow-up at three and six weeks, Alan adult model took less time to make the video; besides,
maintained 100% of both skills. no editing was required. The strategy could be used right
away. Moreover, the teacher would be able to use it with
other students, whereas using a self-model prevented
Social Validity reuse with other students. Although the teacher noted
When participants were asked which video they pre- that using the computer software was relatively “teacher
ferred, all reported that they liked watching themselves. friendly,” she also reported that the technological de-
Moreover, they said that the videos helped them learn mands were greater for making a self-modeling video,
the tasks. Following the study, the classroom teacher and capturing the participant performing the expected
completed the social validity IRP-15 rating scale. The behaviors required greater preparation and time.
teacher rated the video self-modeling an 80 and the adult
modeling condition an 85 – scores well above the 52.5
rating needed for acceptability – suggesting the video
strategy was socially acceptable to use in a school setting Discussion
The teacher rated both video modeling conditions above
The purpose of this study was to examine the effects of
53. The teacher strongly agreed that with both model-
video self-modeling and the use of an unfamiliar adult
ing condition (a) teachers would find the video model-
model on the acquisition and maintenance of a chained
ing procedures acceptable, (b) the strategy was a good
task by adolescents with ASD. Overall, self-modeling
way to teach vocational skills, (c) I liked the procedures
and adult modeling were equally effective and efficient at
used, (d) I would suggest the use of video modeling to
teaching participants chained vocational tasks. In addi-
other teachers, and (e) video modeling did not result in
tion, participants maintained vocational skills three and
negative side effects. The teacher also reported that video
six weeks after acquisition.
modeling would be an easy strategy to use with para-
professional, peer tutors, and substitute teachers to con-
Nevertheless, individual differences emerged. Ryan’s data
vey learning expectations and appropriate assistance for
paths fractionated, indicating a preference for self-mod-
working with individuals with autism.
eling. Two participants, Ronald and Phil, acquired tasks

JSET 2008 Volume 23, Number 3 17

 Journal of Special Education Technology

slightly more efficiently during the self-modeling condi- adding a response-based prompting strategy may be a vi-
tion, while Alan, the remaining participant, showed no able solution when errors persist. Third, the study ex-
preference between self and adult model conditions. Al- amined the social validity of video modeling procedures.
though differences occurred in terms of the mean percent- The classroom teacher rated the use of video modeling
age of task-analyzed steps performed independently, the as highly socially acceptable. The teacher also noted the
number of instructional sessions to reach criterion, and benefits the participants gained from video instruction
the number of errors made, self and adult models were and the ease of using video modeling in the classroom.
functionally analogous for three of the four participants. Furthermore, the teacher reported that the adult model
condition was easier to create and required less time to
These findings confirm that results of previous studies produce; besides, the materials could be reused. Produc-
reporting positive outcomes from using video model- ing self-modeling videos will become more simplified
ing, both self and other models (e.g., Ayres & Langone, and trouble free, decreasing teacher preparation time
2007; Bellini & Akullian, 2007; D’Ateno et al., 2003; with technological advancements.
Ihrig & Wolchok, 1988; Norman, Collins & Schuster,
2001). The study also supports the Sherer et al. (2001) All participants reported that they preferred watching
study, which demonstrated no functional difference be- themselves perform the task, which may explain individ-
tween self and other models for children with ASD ac- ual differences. They also noted that they enjoyed mak-
quiring discrete behaviors. ing the movies, suggesting social validation of the use of
video self-modeling. An advantage of video self-model-
In addition, this study extended the video modeling lit- ing compared to other models includes the possibility
erature in several ways. First, it demonstrated that both of improving self-efficacy. Bray and Kehle (1996) previ-
modeling conditions were equally effective in teaching ously noted the reinforcing value of watching yourself
participants to acquire and maintain vocational chain be successful. According to Bandura (1997), individuals
tasks. Individuals with autism typically have deficits in can acquire self-efficacy through external supports, par-
a wide array of skills that require the display of a chain ticularly through observing their own success. Another
of behaviors to reach mastery (e.g., functional life skills explanation of the self-modeling preference may include
such as dressing, cooking, and bed making; community the familiarity of the self-model. Since the participant
living skills such as shopping and ordering at a restaurant; was the model, increased attention and visual processing
recreation leisure skills such as bowling and operating a of the modeled behaviors may have transpired. Lastly,
DVD player; vocational skills such as food preparation typical adolescents like watching television, and there is
and delivering materials). Video modeling using either no evidence to suggest this is different for adolescents
self or an adult model proved to be a successful strategy with ASD. Besides, the participants in this study were
that enhanced learning for students with ASD. Moreover, known for watching certain television programs and
both conditions improved participants’ initiation to start spending a majority of their free time watching televi-
on a task and the duration of task completion. While not sion and DVDs.
the direct focus of this study, all participants had tran-
sitional goals of beginning and completing job-related Several limitations of this study may have affected the
tasks within similar lengths of time relative to “typical” overall results and interpretations. First, baseline probes
workers. After using video modeling, it was observed were discontinued after 90 s of nonresponsiveness. It is
that all participants started and finished their work more possible, therefore, that participants would have per-
quickly and with greater fluency than before. formed more task-analyzed steps independently given
more time. Second, the study was conducted with par-
Second, previous studies replayed the video when the par- ticipants who generally liked watching television and
ticipant was observed making an error. This study incor- videos. Participants whose preferences are dissimilar may
porated the use of a response-based prompting strategy perform differently using video modeling instruction.
(i.e., system of least prompts) to aid in task acquisition. Third, the adult model was very similar to the partici-
Although having the individual watch the video again pants. That is, the unfamiliar adult was a male in his late
resulted in error performance corrections (Le Grice & twenties of the same ethnicity as the participants. Great-
Blampied, 1994; Tiong, Blampied, & Le Grice, 1992), er model diversity may have produced greater outcome

18 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

variance. Fourth, generalization was not assessed. Gener- Branham, R. S., Collins, B. C., Schuster, J. W., & Kleinert, H. (1999).
alization probes may parcel out whether the functional- Teaching community skills to students with moderate disabilities:
comparing combined techniques of classroom simulation, video-
ity of the model is important. tape modeling, and community-based instruction. Education and
Training in Mental Retardation and Developmental Disabilities, 34,
Results from this study provide support for the use of 170-181.
video self or adult modeling as an effective and efficient Bray, M., & Kehle, T. (1996). Self-modeling as an intervention for
teaching strategy for individuals with ASD. However, stuttering. School Psychology Review, 25, 358-369.
further research is needed to verify these results and to Brown, G., & Middleton, H. (1998). Use of self-as-a-model to pro-
continue exploring various types of models. Future re- mote generalization and maintenance of the reduction of self-
search also should explore the utility of video modeling stimulation in a child with mental retardation. Education and
Training in Mental Retardation and Developmental Disabilities, 33,
for generalization of skills across settings and skills. Mul- 76-80.
tiple videos incorporating several exemplars of various
Buggey, T. (2005). Video self-modeling applications with students
materials in different locations may improve skill gener- with autism spectrum disorder in a small private school setting.
alization. In addition, the use of portable video devices Focus on Autism and Other Developmental Disabilities, 20, 52-63.
should be examined, particularly for use in community
Cannella-Malone, H., Sigafoos, J., O’Reilly, M., de la Cruz, B., Edris-
settings. inha, C., & Lancioni, G. (2006). Comparing video prompting to
video modeling for teaching daily living skills to six adults with
developmental disabilities. Education and Training in Developmen-
tal Disabilities, 41, 344-356.
References Cihak, D., Alberto, P., Taber-Doughty, T., & Gama, R. (2006). A
comparison of static picture prompting and video prompting
Alberto, P., Cihak, D., & Gama, R. (2005). Use of static picture simulation strategies using group instructional procedures. Focus
prompts versus video modeling during simulation instruction. on Autism and Other Developmental Disabilities, 21, 89-99.
Research in Developmental Disabilities, 26, 327-339.
Coyle, C., & Cole, P. (2004). A videotaped self-modeling and self-
Alpern, G., Boll, T., & Shearer, M. (1988). Developmental profile II. monitoring treatment program to decrease off-task behaviour in
Los Angeles: Western Psychological Services. children with autism. Journal of Intellectual & Developmental Dis-
Ayres, K., & Langone, J. (2005) Intervention and instruction with abilities, 29, 3-15.
video for students with autism: A review of the literature. Educa- D’Ateno, P., Mangiapanello, K., & Taylor, B. (2003). Using video
tional and Training in Developmental Disabilities, 40, 183-196. modeling to teach complex play sequences to a preschooler with
Ayres, K., & Langone, J. (2007). A comparison of video modeling autism. Journal of Positive Behavior Interventions, 5, 5-11.
perspective for students with autism. Journal of Special Education Gilliam, J. (1995). Gilliam autism rating scale. Austin, TX: PRO-
Technology, 22(2), 15-30 ED.
Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ; Graves, T. B., Collins, B. C., Schuster, J. W., & Kleinert, H. (2005).
Prentice-Hall. Using video prompting to teach cooking skills to secondary stu-
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: dents with moderate disabilities. Education and Training in Devel-
Freeman. opmental Disabilities, 40, 34-46.

Barlow, D. H., & Hersen, M. (1984). Single case experimental designs: Haring, T. G., Breen, C. G., Weiner, J., Kennedy, C. H., & Bednersh,
Strategies for studying behavior change (2nd ed.). Needham Heights, E. (1995). Using videotape modeling to facilitate generalized pur-
MA: Allyn & Bacon. chasing skills. Journal of Behavioral Education, 5, 29-53.

Bellini, S., & Akullian, J. (2007). A meta-analysis of video model- Harris, S., Handleman, J., & Burton, J. (1990). The Stanford Bi-
ing and video self-modeling interventions for children and ado- net profiles of young children with autism. Special Services in the
lescents with autism spectrum disorder. Exceptional Children, 73, School, 6(1/2), 135-143.
264-287. Hermelin, B., & O’Connor, N. (1970). Psychological experiments with
Bidwell, M. A., & Rehfeldt, R. A. (2004). Using video modeling to autistic children. London: Pergamon.
teach a domestic skill with an embedded social skill to adults with Hodgdon, L. Q. (1995). Solving behavioral problems through the
severe mental retardation. Behavioral Interventions,19, 263-274. use of visual supported communication. In K. A. Quill (Ed.),
Bondy, A., & Frost, L. (1994). The picture exchange communication Teaching children with autism: Strategies to enhance communication
system. Focus on Autistic Behavior, 9, 1-19. and socialization (pp. 265-286). New York: Delmar.
Hosford, R. E. (1981). Self-as-a-model. A cognitive social learning
technique. The Counseling Psychologist, 9, 45-61.

JSET 2008 Volume 23, Number 3 19

 Journal of Special Education Technology

Ihrig, K., & Wolchik, S. A. (1988). Peer versus adult models and Rehfeldt, R. A., Dahman, D., Young, A., Cherry, H., & Davis, P.
autistic children’s learning: Acquisition, generalization, and main- (2003). Teaching a simple meal preparation skill to adults with
tenance. Journal of Autism and Developmental Disabilities, 18, moderate and severe mental retardation using video modeling.
67-79. Behavioral Interventions, 18, 209-218.
Krantz, P. J., & McClannahan, L. E. (1998). Social interactions skills Sherer, M., Perice, K. L., Paredes, S., Kiasacky, K., Ingersoll, B., &
of children with autism: A script-fading procedure for beginning Schreibman, L. (2001). Enhancing conversation skills in children
readers. Journal of Applied Behavior Analysis, 31, 191-202. with autism via video technology: which is better, “self ” or “other”
as a model? Behavior Modification, 25, 140-158.
Kyhl, R., Alper, S., & Sinclair, T. (1999). Acquisition and generaliza-
tion of functional words in community grocery stores using video- Siegel, J., Minshew, N., & Goldstein, G. (1996). Wechsler IQ pro-
taped instruction. Career Development for Exceptional Individuals, files in diagnosis of high-functioning autism. Journal of Autism
22, 55-67. and Developmental Disabilities, 26, 389-406.
Le Grice, B., & Blampied, N. M. (1994). Training pupuls with intel- Sigafoos, J., O’Reilly, M., Cannella, H., Upadhyaya, M., Edrisinha,
lectual disability to operate technology using video prompting. C., Lancioni, G., Hundley, A., Andrews, A., Graver, C., & Young,
Education and Training in Mental Retardation and Developmental D. (2005). Computer-presented video prompting for teaching
Disabilities, 29, 321-330. microwave oven use to three adults with developmental disabili-
ties. Journal of Behavioral Education, 14, 189-201.
Lincoln, A., Courchesne, E., Kilman, B., Elmasian, R., & Allen, M.
(1988). A study of intellectual abilities in high-functioning people Tiong, S. J., Blampied, N. M., & Le Grice, B. (1992). Training com-
with autism. Journal of Autism and Developmental Disabilities, 18, munity-living, intellectually handicapped people in fire safety us-
505-523. ing video prompting. Behaviour Change, 9, 65-72.
MacDuff, G. S., Krantz, P. J., & McClannahan, L. E. (1993). Teach- Van Laarhoven, T., & Van Laarhoven-Myers, T. (2006). Comparison
ing children with autism to use photographic activity schedules: of three video-based instructional procedures for teaching daily
Maintenance and generalization of complex response chains. Jour- living skills to persons with developmental disabilities. Education
nal of Applied Behavior Analysis, 26, 89-97. and Training in Developmental Disabilities, 41, 365-381.
Martens, B. K., Witt, J. C., Elliott, S. N., & Darveaux, D. (1985). Vaughn, B., & Horner, H. (1995). Effects of concrete versus verbal
Teacher judgments concerning the acceptability of school-based choice systems on problematic behaviors. Augmentative and Alter-
interventions. Professional Psychology: Research and Practice, 16, native Communication, 11, 89-92.
191-198.
VonBrock, M. B., & Elliott, S. N. (1987). Influences of treatment
Mechling, L., Gast, D., & Langone, J. (2002). Computer-based video effectiveness information on the acceptability of classroom inter-
instruction to teach persons with moderate intellectual disabilities ventions. Journal of School Psychology, 25, 131-144.
to read grocery aisle signs and locate items. The Journal of Special
Education, 35, 224-240.
Author Notes
Norman, J. M., Collins, B. C., & Schuster, J. W. (2001). Using an in-
structional package including video technology to teach self-help David Cihak, Ph.D., is an assistant professor in the Department
skills to elementary students with mental disabilities. Journal of of Theory and Practice in Teacher Education at the University
Special Education Technology, 16(3), 5-18. of Tennessee at Knoxville. Leigh Schrader, M.S., is a high school
Pierce, K. C., & Schreibman, L. (1994). Teaching daily living skills special education teacher from the University of Tennessee at
to children with autism in unsupervised settings through picto- Knoxville.
rial self-management. Journal of Applied Behavior Analysis, 27,
471-481. Correspondence regarding this article should be addressed to David
Cihak, University of Tennessee, A12 Bailey Education Complex,
Reagon, K., Higbee T., & Endicott, K. (2006). Teaching pretend play
skills to a student with autism using video modeling with a sibling Knoxville, TN 37996-3442. Email to: [email protected]
as model and play partner. Education and Treatment of Children,
29, 517-528.

20 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

The Efficacy of Technology Use by People with Intellectual

Disability:
A Single-Subject Design Meta-Analysis
Michael L. Wehmeyer
Susan B. Palmer
Sean J. Smith
University of Kansas

Daniel K. Davies
Steven Stock
AbleLink Technologies

It is generally assumed that when technology is designed with the principles of universal design in mind,
people with intellectual disability will also benefit. However, there is no research concerning the degree to
which technology use is impacted by the presence or absence of universal design features. To address this and
to examine the degree to which people with intellectual disability benefit from technology, we conducted a
meta-analysis of single subject design studies that implemented technology with a person with intellectual
disability. We found that most of the technology use evaluations with people with intellectual disability have
been with a relatively limited number of technology types and do not include universal design features. The
need for research on the impact of specific types of universal design features is discussed.

T echnology has become a ubiquitous part of life in

the 21st century and to improve one’s quality of life,
one must be able to use technology to some degree. This
et al., 1992, p. 10) and that “mental retardation is a dis-
ability only as a result of this interaction” (p. 10).

is as true for people with intellectual disability as it is for Fundamental to this functional definition was the im-
the general public. In fact, the potential for technology portance of providing supports that meet unique needs.
to contribute to a better quality of life for people with The AAMR manual (Luckasson et al., 1992) defined sup-
intellectual disability is more than simply an issue of con- ports as “resources and strategies that enable people with
venience and becomes one of access. During the 1990s, and without disabilities to access resources, information
there was a movement in the field of intellectual disabil- and relationships inherent within integrated work and
ity toward providing supports instead of programs (Weh- living environments; and that result in their enhanced
meyer & Patton, 2000). The 1992 American Association interdependence, productivity, community integration,
on Mental Retardation (AAMR) definition and classifi- and satisfaction” (p. 101). The shift to a functional defi-
cation system defined “mental retardation” as a state of nition of intellectual disability and the emphasis on sup-
functioning in which limitations in functional capacity ports increases the importance of the role of technology
and adaptive skills must be considered within the context in the lives of people with intellectual disability. Thus,
of environments and supports. The AAMR manual pro- technology provides powerful ways to address limitations
posed that “mental retardation is a state in which func- in functioning and to provide accommodations to en-
tioning is impaired in certain specific ways” (Luckasson vironments in which people with intellectual disability

JSET 2008 Volume 23, Number 3 21

 Journal of Special Education Technology

live, learn, work, or play. Technology use may be impor- 2. Flexibility in use: A design accommodates a wide
tant for most people to ensure a better quality of life; it is range of preferences and abilities;
essential for people with intellectual disability.
3. Simple and intuitive use: Use of the design is easy
And yet, we know little about the degree to which people to understand, regardless of user’s experience,
with intellectual disability use technology. In response to knowledge, language, or cognitive skills;
that gap in information, in 1995 The Arc, a national or-
ganization on intellectual and developmental disabilities, 4. Perceptible information: The design communicates
conducted a national survey of parents/family members the information needed by the user, be it through
regarding technology use by their family member with different modes or by providing adequate contrast;
intellectual disability (Wehmeyer, 1998, 1999). The sur-
vey consisted of five areas of questions focusing on the 5. Tolerance for error: The design minimizes adverse
use of technology for a specific purpose: (a) Mobility consequences or unintended actions;
Technology Devices, (b) Hearing and Vision Technology
Devices, (c) Communication Technology Devices, (d) 6. Low physical effort. The design can be used com-
Home Adaptations, and (e) Environmental Control and fortably with minimum fatigue; and
Independent Living Devices. An additional set of ques-
tions tapped into the use of personal computers. Specifi- 7. Size and space for approach and use appropriate:
cally, the survey solicited information about technology Design allows for approach, reach and manipula-
use in each functional area and computer use and avail- tion and device use independent of user’s body
ability, unmet needs with regard to each of the functional size or mobility. (Trace Center, 1995, p. 1)
areas and computer use, barriers to technology use, train-
ing to use the technology, and satisfaction with technol- It is generally assumed that when technology is designed
ogy use. with the principles of universal design in mind, people
with intellectual disability will benefit. However, to our
Results showed that in all use areas except mobility, the knowledge, no research has studied the degree to which
number of respondents with intellectual disability who technology use is impacted by the presence or absence of
needed technology but did not use it was greater than universal design features. To address this and to examine
the number of respondents who had and used the tech- the degree to which people with intellectual disability
nology. With regard to computer use, generally, people benefit from technology, we conducted a meta-analysis
often had access to computers but rarely used them. The of single-subject design studies that implemented tech-
most frequently cited barrier to technology use was the nology with a person with intellectual disability.
cost or lack of funds, followed by a lack of training, lack
of information about what the computer could do to
benefit the family member, the complexity of the device,
and a lack of assessment. In the years since this survey, Method
the cost of technology devices like personal computers
has decreased, and yet little has been done to address the Procedure
usability factors related to technology.
As part of a federally funded project to examine utiliza-
tion of technology by people with intellectual disability,
These issues of usability are not new to the field of assis-
we conducted an extensive search for articles published
tive technology, which has emphasized universal design
in peer-reviewed journals on the use of technology by
for decades. The Trace Center in Wisconsin identified
people with intellectual disability. Specifically, an exten-
seven principles of universal design to consider when de-
sive search of the PsychINFO and ERIC databases for
signing AT:
articles published from 1977 to 2003 was conducted us-
1. Equitable use: A design is useful and marketable to ing two key words (mental retardation, intellectual dis-
any group of users; ability) with a combination of the following other key
words: computer, computer use, technology, technology
use, assistive technology, adaptive technology, instruc-

22 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

tional technology, electronic technology, information Castro, 1987). PND scores can range from 0 to 100%,
technology, assistive devices, augmentative communica- with higher scores indicating more effective treatments.
tion, and alternative communication. Scotti et al. (1991) suggested specific criteria for evaluat-
ing the practical implication of PND values, with PND
We were interested in cataloging all articles meeting the scores ranging from 50% to 80% identifying the treat-
simple key word search criteria. Thus, initial information ment as questionable, scores greater than 80% and less
(article title, authors names, journal name, volume num- than 99% as fair, and scores greater than 99% as highly
ber, and page numbers) was entered into a Microsoft effective. Baseline data reaching a floor level of perfor-
Access database. Each article was then obtained and re- mance (i.e., zero) will result in a PND of zero, which
viewed to gain more information. A total of 411 articles in some circumstances may not be an appropriate repre-
that fit the search criteria were retrieved and information sentation of treatment effects. For this reason, Scruggs,
from each was coded. Further, information added to the Mastropieri, & Castro, (1987) hold that when not more
database for these 411 articles included keywords for each than 3, or less than 33-1/3% of total baseline data points
article, information on the type of technology used, the reach a floor level, PND cannot be calculated.
functional use area to which the technology was applied
(communication, mobility, environmental control, daily PZD, in turn, is a companion measure of the degree to
living, community inclusion, employment, education, which an intervention reduces and maintains a behavior
recreation/leisure), level and type of cognitive or physical at zero levels. It is calculated by identifying the first data
impairment, and research design. We also evaluated the point in the treatment phase that reached zero and cal-
degree to which any issue of universal design (equitable culating the percentage of data points from that point
use, flexible use, simple and intuitive use, perceptible onward that remained at zero level (Scotti et al., 1991).
information, tolerance for error, low physical/cognitive Percentage zero data scores also range from 0 to 100%
effort, size and space) was discussed or identified as fea- with higher scores indicating more effective treatments.
tures of the device evaluated in a given study. Scotti et al. (1991) suggested specific criteria for evaluat-
ing the practical implication of PZD values, with PZD
Of the 411 articles, 275 were databased; the remaining scores under 18% identifying the treatment as ineffec-
136 were opinion articles or position statements. Of the tive, scores from 18% to 54% as questionable, scores
275 databased articles, 251 were quantitative (group de- from 55% to 80% as fairly effective, and any score over
sign, single-subject design, literature reviews), and 24 80% as highly effective.
were qualitative. Of the quantitative studies, 81 imple-
mented a single-subject design.1 It was with this group The 81 single-subject design studies selected for inclusion
that we conducted the single-subject design meta-anal- in the meta-analysis involved a total of 281 participants
ysis. Among the small-subject designs implemented in with intellectual and developmental disability. However,
these studies, two-thirds were multiple-baseline design in many of the studies, a single subject was involved in
studies (n = 54), 15% used an ABAB reversal design (n = multiple tasks, leading to multiple graphs. As such, each
12), 10% used an AB design (n = 8), and the remaining task/graph was counted as a unique case, and PND/PZD
studies used some other design. analyses were conducted with regard to those cases (n
= 456). The 281 participants ranged in age from 2 to
Each of the 81 studies was examined for treatment ef- 68 years (mean age = 15.96; SD = 8.61). Articles were
ficacy to calculate the percentage of nonoverlapping data coded to record each participant’s age, gender, diagnosis,
(PND) (Scruggs & Mastopieri, 1998; Scruggs, Mas- setting, and, when available, IQ score. IQ scores were
topieri, & Casto, 1987) and percentage zero data (PZD) reported for only 106 of the 281 participants and ranged
(Scotti, Evans, Meyer, & Walker, 1991) metrics as indices from 15 to 75 (mean IQ score = 45.67; SD = 13.10). Par-
of behavior change. PND is a measure of the proportion ticipants consisted of 169 males and 106 females (gen-
of nonoverlapping data between baseline and treatment der data were missing for 6 participants). Males ranged
phases. It is calculated by dividing the number of treat- in age from 2 to 68 (mean age = 17.23; SD = 8.83),
ment data points that fall below the lowest baseline data and IQ scores (available for 53 male participants) ranged
point by the total number of data points in the treat- from 15 to 75 (mean IQ score = 46.36; SD = 13.47).
ment phase, multiplied by 100 (Scruggs, Mastropieri, & Females ranged in age from 8 to 41 (mean age = 17.52;

JSET 2008 Volume 23, Number 3 23

 Journal of Special Education Technology

SD = 8.31), and IQ scores (available for 47 female par- percent of total number of cases (n = 456) are reported
ticipants) ranged from 18 to 75 (mean IQ score = 45.87; in Table 1.
SD = 13.34).
To assist in subsequent analyses by technology type, we
collapsed the 19 technology types into 9: (a) comput-
ers; (b) augmentative or alternative communication de-
Analyses vices; (c) electronic and information technology (e.g.,
cell phones); (d) auditory prompting systems; (e) video-
PND and PZD scores were calculated for each unique
based systems (TV, interactive disk, etc.); (f ) palmtop or
treatment phase and its preceding baseline identified in
hand-held computers; (g) switches; (h) voice-recognition
the cases (n = 456). For studies that utilized an ABAB
software; and (i) home appliances.
design with one or more participants, only treatment
phases with a baseline preceding them were analyzed,
With regard to the use of the technology, we identified six
but each baseline and treatment phase was treated as
overarching use categories: (a) academics; (b) computer
a unique case. Data were summarized and reported in
skills/use training; (c) community inclusion, including
tabular and graphic formats. To examine the effect of in-
community-based living, leisure and employment uses;
dividual characteristics and universal design features on
study outcomes, separate ANOVAs were conducted with
PND scores as dependent variables.

The reliability of PND scores was assessed by hav- Table 1

ing two raters independently calculate PND scores for
each unique treatment phase and its preceding baseline. Technology Type by Study Participant
Agreement was scored when the first and second coders Technology Type Frequency %
obtained identical PND and PZD scores for each unique
phase. Reliability was calculated by dividing the number Computer–Educational 123 27.0
of agreements by the number of agreements plus dis- Augmentative Communication 74 16.2
agreements, multiplied by 100. Coders agreed complete- Auditory Prompting Device 54 11.8
ly (100%) for eight of the single-subject design articles, Video Device 45 9.9
with the remaining articles reaching 90% agreement or Switches 36 7.9
above. The coders came to consensus on the remaining
Video Tape 23 5.0
data point computations to ensure the most accurate
Electronic Technology 22 4.8
scores were entered into the analysis
Computer–Recreational 12 2.6
Computer-Assisted Electronics 12 2.6
Palmtop Computer–Audio
Results Vibrator 10 2.2
Visual Orientation Robot 8 1.8
Characteristics of Studies Microswitch and Tape
Recorder 8 1.8
Of the 81 studies, 38 (47%) involved an initial test or
evaluation of a device; 11 (14%) were not initial eval- Voice-Recognition Software–
Environment Control 6 1.3
uations of a device or technology, but were the initial
application to the population of people with intellec- Appliances 6 1.3
tual disability; 16 (20%) involved comparisons of two Palmtop Computer 4 .9
or more technologies or technology methods; and 16 Television 4 .9
(20%) involved the application of a previously evaluated Microphone 4 .9
technology to a different dependent variable. As to type Interactive Video Disk 3 .7
of technology used, 19 different technology types were Optic Sensors 2 .4
reported. These types and their relative frequency and

24 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

(d) communication; (e) self-regulation; and (f ) mobility/ peared. Finally, with regard to descriptions of the studies
environmental control. Table 2 provides frequency (by themselves, Table 3 lists the settings in which the tech-
study) and % of total studies in which these uses ap- nology was implemented. Because a given device might

Table 2

Technology Use Categories by Study

Use Category n % Avg PND

Academic
Mathematics 3 3.7 .56
Language Arts 12 14.8 .94
Computer Skills/Computer Use 3 3.7 .96
Community Living/Leisure, Employment 20 24.7 .84
Communication 19 23.5 .74
Self-Regulation 16 19.8 .94
Mobility/Environmental Control 8 9.9 .86

Table 3

Setting of Intervention by Participants

Setting Freq % Avg. PND

Special Education Setting

Academic 207 45.4 .71

Vocational 32 7.0 .89
Functional 3 .7 1.00
General 9 2.0 .81
University 6 1.3 .98
Residential Setting
Parental Home 7 1.5 .83
Group Home 1 .2 1.00
Congregate Setting 90 19.7 .89
Hospital 6 1.3 .33
Employment Setting 8 1.8 .75
Community Setting, General 57 12.5 .87
Computer Lab 17 3.7 .89
Multiple Settings 5 1.1 .85
Other Settings 8 1.8 1.00

JSET 2008 Volume 23, Number 3 25

 Journal of Special Education Technology

be implemented in more than one setting in any one

Table 4
study, these data are reported by case (n = 456)
PND Scores by Technology Group
Percent Non Overlapping and Percent Zero
Data Technology Group Mean PND Score (SD)
Palmtop Computer 0.99 (.002)
The mean percentage nonoverlapping data (PND) score
Auditory Prompt/Device 0.94 (.10)
for all 456 cases was 80% (SD = .29), indicating that
the use of technology by persons with intellectual dis- Electronic and Information
Technology 0.96 (.16)
ability resulted in Fair effects. However, PND scores var-
ied by technology group, with PND scores for palmtop Video Device (TV, interac-
tive disk) 0.82 (.23)
computers indicating High effectiveness (99%). For au-
ditory prompting devices, electronic devices, video de- Augmentative
Communication 0.81 (.27)
vices, and augmentative communication devices scores
fell between 80% and 99%, thus, indicating Fair efficacy. Computer 0.71 (.33)
The remaining technology groups (computer, home ap- Home Appliances 0.70 (.42)
pliances, voice recognition, switches) all showed PND Voice Recognition Software
scores below 80% and fell in the Questionable efficacy for Environmental Control 0.67 (.17)
range. As shown in Table 2, when examined by technol- Switches 0.60 (.37)
ogy use, all PND scores fell in the Fair range, with PND
mean scores ranging from 84% to 96%. The only excep-
tions were two use areas, Academics Mathematics and
Communication, the PND for both of which was in the
With regard to the PZD statistic, only 27 cases had
Questionable range (communication devices, this was
PZD scores. The PZD metric is a measure of how well
just a few percentage points away from the Fair range).
an intervention to reduce target behavior works, with
Similarly, as shown in Table 3, when examined by inter-
the assumption that no instances of such behaviors is
vention setting, most PND scores fell in the Fair range.
preferable. This almost universally applies to attempts
Several settings obtained PND scores of 100%, though
to decrease problem behavior, and there were not many
those were typically low frequency settings. Employment
studies in the sample intending to reduce behavior. The
and Special Education-Academic settings fell out of the
mean PZD score for the 27 cases was 78% (SD = .27),
Fair range and into the Questionable range, and PND
which placed this in the Fairly Effective range.
scores from the Hospital setting fell out of the Question-
able range. Table 4 provides PND scores by technology
group. Application of Universal Design Features
In all, 183 of the 456 (40%) cases studied used a de-
Significant differences on PND scores emerged by level
vice that had at least one universal design (UD) feature
of severity of intellectual disability (mild, moderate, se-
identified, and 273 (60%) used a device for which no
vere, profound) using ANOVA, F(3, 452) = 16.29, p =
UD feature was identified. Table 6 presents these data
.0001. Post hoc analyses (Scheffé tests) indicated differ-
broken down by UD principle. Because more than one
ences between (a) cases in which participants had mild
UD feature could be present for any device, the pres-
intellectual disability and cases in which participants had
ence or absence of an identified UD feature was noted by
moderate or severe intellectual impairment, (b) cases in
each UD principle. Finally, an ANOVA of PND scores
which participants had moderate intellectual disabil-
by presence or absence of a UD feature revealed signifi-
ity and profound intellectual disability, and (c) cases in
cant differences between these two groups (e.g., cases in
which participants had severe impairment and profound
which at least one UD feature was identified and cases in
intellectual disability. No significant differences were
which no UD feature was identified), F(1, 454) = 14.46,
found in PND scores by gender. Table 5 presents PND
p = .0001. Participants in the group incorporating UD
mean differences, standard error, and statistical signifi-
features had an average PND score of 86% (SD = .24)
cance data from the post hoc analysis.

26 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

and those in the group not addressing UD features had Discussion

an average PND score of 75% (SD = .31).
The results from this meta-analysis provide direction for
future research on technology use by people with intel-
lectual disability. Before considering these findings, it is
worth noting the limitations to the study, however. First,

Table 5

Post Hoc Analysis of PND Scores by Severity of Disability

PND Mean Std.

Disability Severity of Disability Difference Error Sig.
Mild Intellectual Disability Moderate Intellectual Disability -.19* .04 .000
Severe Intellectual Disability -.19* .04 .000
Profound Intellectual Disability .007 .05 .999
*
Moderate Intellectual Disability Mild Intellectual Disability .19 .04 .000
Severe Intellectual Disability .004 .03 1.000
Profound Intellectual Disability .20* .04 .000
*
Severe Intellectual Disability Mild Intellectual Disability .19 .04 .000
Moderate Intellectual Disability -.004 .03 1.000
Profound Intellectual Disability .196* .04 .000
Profound Intellectual Disability Mild Intellectual Disability -.007 .05 .999
*
Moderate Intellectual Disability -.20 .04 .000
*
Severe Intellectual Disability -.20 .04 .000
* The PND mean difference is significant at the 0.05 level.

Table 6

Application of Universal Design Principles

UD Feature No UD Feature % Cases UD Feature % Cases No UD

UD Principle Identified Identified Identified Feature Identified
Equitable Use 36 420 7.9 92.1
Flexible Use 121 335 26.5 73.5
Simple Intuitive Use 35 421 7.7 92.3
Perceptible Information 15 441 3.3 96.7
Tolerance for Error 26 430 5.3 94.3
Low Phys/Cog Effort 4 452 .9 99.1
Size and Space 0 456 0.0 100.0

JSET 2008 Volume 23, Number 3 27

 Journal of Special Education Technology

we made no effort to screen out studies by quality of of principles of UD, or the lack thereof, can account for
the research design. Findings, whether positive or nega- these issues, although there is no mechanism to attribute
tive, are influenced by the control established in small- causality inherent in this study design. The ANOVA re-
subject design, by the number of subjects in the study sults showing that people using universally designed de-
and by other factors that impact the power of the study. vices had higher PND scores supports this hypothesis,
As such, the PND scores must be judged based upon but there were too few studies by purpose to evaluate
this limitation. Second, we relied exclusively on the ar- this for specific areas such as communication. Another
ticles reviewed to identify if a person with intellectual viable explanation for this finding is that people who
disability was involved. As a result, we have limited data need communication devices may have more severe cog-
on such factors as IQ (available for only 40% of the par- nitive impairments, thus limited efficacy data would be
ticipants). Perhaps more important, the limited data on as attributable to the population examined as the device
IQ and severity of disability limited what we could say category. Because of the limited availability of IQ data,
about the impact of level of intellectual impairment on we opted not to evaluate this area, and, therefore, it re-
technology use to the global finding that people with the mains an area of research need.
most severe disability had the least positive PND scores.
Third, too few studies reported the impact of technology With regard to setting, although PND scores generally
on reducing behavior to allow us to make more specific ranged in the Fairly Effective category, perhaps the most
statement about the PZD scores than that the outcomes striking outcome was that almost three quarters of the
were promising. participants were evaluated in a substantially segregated
setting. Specifically, 53% were evaluated in a special edu-
Given these caveats, the first observation to be made is cation setting (i.e., outside the general education setting),
that, as evidenced in Table 1, most of the technology whereas just over 21% of participants were evaluated
use evaluations with people with intellectual disability in congregate living arrangements. We must conclude,
have involved a relatively limited number of technology therefore, that we know too little about the impact of the
types. Thus, over 55% of participants were evaluated use of technology in typical, community-based settings.
with technology that fell into three types: computers in Take the area of employment as an example. In a sepa-
education, augmentative communication devices, and rate study, we examined the use of technology by people
auditory prompting devices. If the next highest types, with intellectual disability for employment purposes and
video devices and switches, are added, more than 70% of found only 13 single-subject design studies addressing
participants are accounted for. Based on these findings, employment issues, but the technology use was fairly ef-
it is evident that there is a need for research on the ef- fective (Wehmeyer et al., 2006). In the present study, we
ficacy of a wider range of technology devices with people combined employment-related studies with other studies
with intellectual disability. For special educators, it is en- looking at other use areas (independent living, leisure)
couraging that 45% of the participants were evaluated because there were not enough of these use areas to be
using computer technology for educational purposes. looked at separately. Again, technology use was Fairly
However, subsequent findings about the efficacy of such Effective in this broader category, inclusive of employ-
studies and the settings in which they were conducted ment (84%). However, when one examines the percent-
mitigate this positive finding to some degree. age and frequency of participants evaluated within actual
employment settings (1.8%, n = 8 participants), it is evi-
The good news, perhaps, is that technology use has gen- dent that we know very little about technology use and
erally been found to be fairly effective, whether viewed employment issues for people with intellectual disability
by study or participant, or by category, setting, or group. in real employment settings.
The overall mean PND was 80%, indicating that tech-
nology as a whole was Fairly Effective. Nevertheless, there Palmtop computers were the only technology group for
are exceptions to this overall conclusion. When examin- which the PND score rose to the level of Highly Effective
ing the PND of technology use by category, for example, (PND = 99%). While the limited number of studies in
technology used for communication (n=19 studies) pur- this area commend caution in generalizing this finding too
poses had a PND of 74%, placing it in the Questionable broadly, several aspects of palmtop computer technology
range of efficacy. Subsequent issues pertaining to the use might contribute to this finding. Most important, these

28 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

devices are relatively new and have been designed with also were generally effective, a finding that is somewhat
or already embody multiple aspects of UD. For example, surprising considering the lack of UD features identified
the PND for typical computer use was .71, placing it in for the various technology types and devices, as depicted
the Questionable range. Desktop computers and palm- in Table 6. It must be concluded that the bulk of the
top computers differ in some features that impact cogni- literature has evaluated technology that contains few or
tive access. For example, all palmtop devices operate with minimal use of UD features. Only “Flexible Use” was re-
a touchscreen input, and while touchscreens are available ported in a significant number of studies (26.5%), with
for desktop computers, they tend not to have been used every other design feature identified in fewer than 10%
in the studies reported. Instead, input for desktop com- of studies. While there may have been features inherent
puters tends to rely on standard input devices, primarily in devices that were not explicitly identified in the ar-
keyboards and computer mice, neither of which are par- ticles, given the total lack of information on such devices
ticularly intuitive or user friendly. Because of the wide- and the tendency for developers not to consider cogni-
spread use of palmtop computers as prompting devices, tive accessibility issues, we suggest these findings are rep-
an issue discussed subsequently, much of the existing resentative of the state of the field with regard to UD and
software for these devices has been designed to be cog- cognitive access. Flexible Use is an important feature for
nitively accessible. That is important. Because palmtops cognitive access, and it is encouraging that many stud-
are relatively inexpensive and ubiquitous, operate using ies identified that. However, design features emphasizing
simple input devices, are portable, and have audio and Simple and Intuitive Use and Tolerance for Error may be
video output and audio input capacity, they have been as important, or more so, for cognitive access, and these
identified as having potential benefit for people with in- features appear not to be imbedded in most technologies
tellectual disability, and software programs designed to evaluated with people with intellectual disability.
run on these devices, including modified desktops, have
been designed with features of UD for cognitive access in Overall, we would conclude from this study that the glass
mind (Wehmeyer, Smith, & Davies, 2005; Wehmeyer, is half full, as opposed to half empty, with regard to the
Smith, Palmer, Davies, & Stock, 2004). potential efficacy of technology use for people with in-
tellectual disability. For despite the fact that few devices
The fact that the second highest PND by group was audi- studied seem to embody principles of UD that would
tory prompting devices adds another level of explanatory promote cognitive access, interventions were still fairly
power to the efficacy of palmtop computers. That is, the effective. Further, devices with many cognitive access
use of technology for prompting people with cognitive features, like palmtop computers, seem highly effective.
disabilities, which is often a use for palmtop devices, is There is a clear need for more research and development
a simple and effective use of technology. Several studies on a wider range of technologies, including newer elec-
have found that the use of prompting devices reduces the tronic and information technologies, and for determin-
need for and use of external prompts from a person (We- ing the impact of particular UD features on technology
hmeyer et al., 2004). For example, Riffel et al. (2005) ex- usability and efficacy for individuals with intellectual dis-
amined the use of a palmtop prompting device in assisting ability and to do so in typical, inclusive settings instead
four adolescents with intellectual disability to complete of substantially segregated settings, as the case has been
transition-related tasks more independently. Three of the to date. Technology has considerable potential, for ex-
four students (one student did not complete the study) ample, to ensure access to the general education curricu-
significantly reduced the number of prompts from a job lum for students with intellectual disability (Wehmeyer,
coach needed to independently complete a vocational/ Smith, & Davies, 2005), and research in this area should
transition task with the use of the device. Specifically, one be conducted in general education classrooms. Further,
student went from an average of more than 300 prompts technology to promote employment must be conducted
needed to complete the task during baseline to an aver- in actual work settings, technology to promote com-
age of about 40 using the device. The use of technology munity inclusion in typical community settings, and so
to prompt students is effective and important. forth. The promise of the impact of such technology uses
is significant; but until we move forward, it remains sim-
Other technology groups, including video-based and oth- ply a promise.
er electronic and augmentative communication devices,

JSET 2008 Volume 23, Number 3 29

 Journal of Special Education Technology

References Wehmeyer, M. L., Smith, S. J., Palmer, S. B., Davies, D. K., & Stock,
S. (2004). Technology use and people with mental retardation.
Luckasson, R., Coulter, D. L., Polloway, E. A., Reiss, S., Schalock, R. In L. M. Glidden (Ed.), International review of research in men-
L., Snell, M. E., Spitalnick, D. M., & Stark, J. A. (1992). Mental tal retardation (Vol. 29, pp. 293-337). San Diego, CA: Academic
retardation: Definition, classification, and systems of supports. Wash- Press.
ington, DC: American Association on Mental Retardation.
Riffel, L. A., Wehmeyer, M. L., Turnbull, A. P., Lattimore, J., Davies, Author Notes
D., Stock, S., & Fisher, S. (2005). Promoting independent perfor-
mance of transition-related tasks using a palmtop PC-based self- Michael L. Wehmeyer is a professor, Department of Special Educa-
directed visual and auditory prompting system. Journal of Special tion, director of the Kansas University Center on Developmental
Education Technology, 20(2), 5-14. Disabilities, and senior scientist, Beach Center on Disability, The
Scotti, J. R., Evans, I. M., Meyer, L. H., & Walker, P. (1991). A University of Kansas. Susan B. Palmer is an associate research pro-
meta-analysis of intervention research with problem behavior: fessor, Beach Center on Disability, The University of Kansas. Sean
Treatment validity and standards of practice. American Journal on J. Smith is an associate professor of special education, Department
Mental Retardation, 96(3), 233-256. of Special Education, The University of Kansas. Daniel K. Davies
Scruggs, T. E., & Mastropieri, M. A. (1998). Summarizing single- is the president, AbleLink Technologies, Colorado Springs, Colo-
subject research: Issues and applications. Behavior Modification, rado. Steven Stock is a vice president, AbleLink Technologies.
22(3), 221-242.
Correspondence should be addressed to Michael L. Wehmeyer,
Scruggs, T. E., Mastropieri, M. A., & Casto, G. (1987). The quantita-
tive synthesis of single-subject research: Methodology and valida-
Beach Center on Disability, 1200 Sunnyside Avenue, Room 3136,
tion. Remedial and Special Education, 8(2), 24-33. Lawrence, Kansas 66045. Email to [email protected]
Trace Center. (1995). Principles of universal design. Madison, WI: Support for this study was provided by the U.S. Department of
University of Wisconsin – Madison, Trace Center Education’s National Institute for Disability Rehabilitation Re-
Wehmeyer, M. L. (1998). A national survey of the use of assistive search (NIDRR) through grants #H133A01602 (Disability Re-
technology by adults with mental retardation. Mental Retardation, habilitation Research Project on Technology Use by People with
36, 44-51. Mental Retardation) awarded to the University of Kansas, and
Wehmeyer, M. L. (1999). Assistive technology and students with #H133E040019 (The RERC on Advancing Cognitive Technolo-
mental retardation: Utilization and barriers. Journal of Special gies) awarded to the University of Colorado and the University
Education Technology, 14(1), 48-58. of Kansas, with additional funding provided by the Coleman In-
Wehmeyer, M. L., Palmer, S., Smith, S., Parent, W., Davies, D., &
stitute for Cognitive Disabilities at the University of Colorado.
Stock, S. (2006). Technology use by people with intellectual and Endorsement by the federal government should not, however, be
developmental disabilities to support employment activities: A inferred. The authors would like to acknowledge the efforts of
single-subject design meta analysis. Journal of Vocational Rehabili- Mary Boatright and Shelly Gigous in data entry and coding.
tation, 24, 81-86.
Footnotes
Wehmeyer, M. L., & Patton, J. (2000). Mental retardation in the 21st
1
Century. Austin, TX: PROED. A list of the articles included in the review can be obtained from
Wehmeyer, M. L., Smith, S., & Davies, D. (2005). Technology use
the first author.
and students with intellectual disability: Universal design for
all students. In D. Edyburn, K. Higgins, & R. Boone (Eds.),
Handbook of special education technology research and practice (pp.
309-323). Whitefish Bay, WI: Knowledge by Design.

30 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Comparison of Static Picture and Video Prompting

on the Performance of Cooking-Related Tasks by
Students with Autism
Linda C. Mechling
Melissa R.Gustafson
University of North Carolina–Wilmington

This study compared the effectiveness of static photographs and video prompts on the independent task
performance of six young men with a diagnosis of autism. An adapted alternating-treatment design with
baseline, comparison, withdrawal, and final treatment conditions was used to measure the percentage of
cooking-related tasks completed independently by each student across the two treatments. Both procedures
were effective in increasing correct task performance for all participants from baseline levels where only
verbal task directions were provided. However, participants independently completed a greater number of
tasks when using video prompting. During the final treatment phase each student further increased his level
of performance when using video prompting with sets of tasks receiving static picture prompts during the
comparison phase.

P ractitioners working with students with autism spec-

trum disorders (ASD) find identification of effective
teaching methods to be especially demanding (Simpson,
Quill (1997) categorized visually cued instruction as (a)
instructional prompts–visual cues that accompany gestural
and verbal prompts or (b) environmental prompts–visual
2005). Because no single intervention is consistently ef- cues available to students without the presence of other
fective with every student with ASD (Higgins & Boone, prompts. Agran, Fodor-Davis, Moore, and Deer (1989)
1996), teachers and therapists must keep abreast of a refer to environmental prompts as externally generated
range of research-based strategies and align instructional mediators that are prearranged visual or auditory (audi-
methods with the individual needs of the student (Simp- tory tapes) cues that prompt a behavior. These prompts
son, 2005). transfer stimulus control back to the student (Wilson,
Schepis, & Mason-Main, 1987) and decrease reliance on
Students with ASD traditionally demonstrate impair- other persons for assistance. Visual prompting systems are
ments in the areas of attention, information processing, primarily used to prompt completion of complex tasks
and memory (Quill, 1997). With regard to attention, with multiple steps (making a sandwich), change tasks
they demonstrate difficulty rapidly shifting attention, independently (transitioning), or complete a sequence
shifting attention between visual and auditory stimuli, of activities (morning routine) (MacDuff, Krantz, &
and attending to the most important and salient features McClannahan, 1993). Picture prompts can be accessed
of stimuli (Quill). Strengths lie in sustained attention through charts and notebooks, while video prompts and
and visual perception (Quill), which make persons with models can be made accessible using portable DVD play-
ASD candidates for learning through pictures and video, ers, laptop computers, or hand-held devices.
which allow information to be processed through con-
crete visual representations of tasks and directions. Pictures have frequently been evaluated as a form of visu-
al supports for persons with ASD (Lancioni & O’Reilly,

JSET 2008 Volume 23, Number 3 31

 Journal of Special Education Technology

2001; Mechling, 2007). According to Lancioni and 1999); and social communication (Apple, Billingsley, &
O’Reilly, picture cues can be divided into two subgroups: Schwartz, 2005; Thiemann & Goldstein, 2001). Efforts
those that target multi-step tasks and those that sequence have also recently been made to evaluate video technol-
activities through a schedule (activity schedules). ogy as a means of presenting activity schedules to persons
with ASD (Dauphin, Kinney, & Stromer; 2004; Stromer,
Students with ASD have successfully used pictures to Kimball, Kinney, & Taylor, 2006).
complete daily living skills (e.g., setting a table, mak-
ing a bed) (Pierce & Schreibman, 1994); facilitate In summary, static pictures and video technology are
transitions at home and in the community (Dettmer, supported in the literature as effective means of provid-
Simpson, Myles, & Ganz, 2000); improve on-task and ing assistance and instructional cues to persons with
on-schedule behaviors (Bryan & Gast, 2000; Hall, Mc- ASD. The question remains whether one system or form
Clannahan, & Krantz, 1995; Massey & Wheeler, 2000); of prompting is more effective than others for some
self-initiate leisure activities at home (MacDuff, et al., learners. While limited in number, some efforts to ad-
1993); complete transitions at school (Schmit, Alper, dress this question have been made. For example, work-
Raschke, & Ryndak, 2000); increase functional commu- ing with students with intellectual disabilities, Stephens
nication (Charlop-Christy, Carpenter, Le, LeBlanc, & and Ludy (1975) found that motion picture sequences
Kellet, 2002; Schwartz, Garfinkle, & Bauer, 1998; Tin- produced higher posttest scores than photographic slides
cani, 2004); and sustain and independently participate and live models in teaching action concepts (i.e., jump-
in activities at home (Krantz, MacDuff, & McClanna- ing, opening a door, filling a glass of water). Two related
han, 1993). studies (Alberto, et al., 2005; Cihak, et al., 2006) found
no significant difference in effectiveness or efficiency be-
As another form of visual support, Sturmey (2003) tween static pictures and video modeling within a one-
described video technology as a novel and expanding to-one (Alberto, Cihak & Gama, 2005) or small-group
technology for providing positive behavior support to arrangement (Cihak, Alberto, Taber-Doughty, & Gama,
individuals with disabilities. Video-based technology has 2006) when teaching purchasing using a debit machine
been used to support acquisition of a range of academic, and banking using ATM machine.
social, communication, and daily living skills for students
with ASD (Ayres & Langone, 2005; Bellini & Akullian, The two studies used an intervention package that in-
2007; Delano, 2007; Mechling, 2005). Targeted skills cluded community-based instruction, a least-to-most
have included self-initiation of social interactions with prompting procedure, and picture or video prompts.
peers (Nilopoulos, & Keenan, 2003, 2004; Wert & Therefore, the effects of static pictures and video cannot
Niesworth, 2003); purchasing (Alcantara, 1994; Haring, be isolated to determine which components of the treat-
Kennedy, Adams, & Pitts-Conway, 1987); perspective ment packages were responsible for the reported results.
taking (Charlop-Christy & Daneshvar, 2003); imagi- The researchers, in both studies, suggest that the ben-
native play (D’Ateno, Mangiapanello, & Taylor, 2003); efits of static pictures include focus on relevant stimuli,
generative spelling (Kinney, Vedora, & Stromer, 2003); minimization of environmental distractions, and lack of
self-help skills (Hagiwara & Myles, 1999); pretend play motion that may divert attention. The researchers fur-
(Reagon, Higbee, & Endicott, 2006); daily living skills ther note that when applied with students with minimal
(Shipley-Benamou, Lutzker, & Taubman, 2002); and community-based instruction, picture prompts may be a
reduction of tantrums and aggression accompanied by better teaching strategy than video modeling for students
increased language skills and social interactions (Buggey, whose characteristics include high distractibility, difficul-
2005). ty maintaining attention, and poor discrimination.

Communication skills have also been the focus of several In contrast Mechling and Gustafson (in press) found
studies evaluating video technology (Charlop-Christy, video to be more effective than static pictures in prompt-
Le, & Freeman, 2000; Charlop & Milstein, 1989). Of ing six students with moderate intellectual disabilities
special interest have been: verbal responses to questions to complete cooking-related tasks. It appears from the
(Buggey, Toombs, Gardener, & Cervetti, 1999); play- results of the study that video images presented on a
related comments to siblings (Taylor, Levin, & Jasper, portable DVD player presented more information to

32 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

participants, enabling them to independently complete depicting skills or (b) video segments demonstrating the
a greater percent of tasks. These results support those of skill.
Gepner, Deruelle, and Grynfeltt (2001), who compared
the ability of young children with ASD and children
without disabilities to process facial expressions using
motion (videotapes), still videotapes, and photographs. Method
The study found no significant difference or “worse per-
formance” by the children with ASD compared to the Participants
control group, regardless of the condition. In addition,
Six male high school-age students with a diagnosis of
Gepner et al. reported that video recordings used in the
ASD participated in the study (see Table 1). All students
study (a) induced immediate imitation of facial expres-
were imitative and had a prior history of using photo-
sions, (b) maintained students’ high level of focus, and
graphs to complete tasks or follow activity schedules. All
(c) may have created video facial expressions that were
students also had experience watching video recordings
perceived as more natural and valid by the children with
for entertainment and could attend to a video tape for
ASD than those presented on photographs.
one minute. Students were screened for the following ad-
ditional skills: (a) fine-motor ability to operate kitchen
Given the limited amount of research and mixed results
appliances and gadgets and to open boxes, containers,
of comparing the effects of static pictures and video
and bags; and (b) ability to imitate a video model.
technology, the purpose of the current study was to sys-
tematically replicate the study conducted by Mechling
and Gustafson (in press) with students with ASD. The Tasks, Settings, Materials, and Equipment
primary objective of the study was to compare the abil-
Twenty cooking-related tasks (10 tasks per set) identical
ity of students with ASD to complete cooking-related
to those in the Mechling and Gustafson (in press) study
motor tasks when presented with (a) single static pictures
were used. The original tasks (Mechling & Gustafson)
were selected based on sampling a range of kitchen stim-

Table 1

Student Characteristics, Treatment, and Tasks

Vineland
Adaptive
Behavior Static
Age (Year- Autism ID Age at Diagnosis Composite Picture Video
Student Month) Diagnoses (Year-Month) IQ CARSa Score Prompting Prompting
Group 1
Gary 17-8 Moderate 3-4 64c 38 64 Set 1 Set 2
b
Eddie 17-10 Mild 3-3 49 31 44 Set 1 Set 2
b
Dan 21-1 Moderate 3-2 44 UA 58 Set 1 Set 2
Group 2
Jessie 16-3 Moderate 6 yr 40d 34.5 38 Set 2 Set 1
c
Kyle 15-10 Moderate 3-4 40 UA 53 Set 2 Set 1
c
Josh 16 yrs 8m Moderate 4 yrs 54 35 64 Set 2 Set 1

Note: aChildhood Autism Rating Scale; bKaufman Assessment Battery for Children; cWechsler Intelligence for Children (3rd
ed); dLeiter International Performance Scale Revised (Leiter-R). UA = unavailable.

JSET 2008 Volume 23, Number 3 33

 Journal of Special Education Technology

uli and response requirements and due to their inclu- Video prompting. A separate video segment was creat-
sion in recipes from three picture cookbooks designed ed for each cooking-related task using a SONY digital
for non-readers (Look’n Cook Microwave: Easy-to-Make DVD-RW 1.4 GB video camera. Video models of each
Illustrated Recipes, Hanson, 1999; Stepwise Lunch Cook- task were recorded directly onto DVD-RW, 30-minute
book, Jackson, 1998; Visual Recipes: A Cookbook for Non- discs. A TOSHIBA portable DVD player with a 7 in.
Readers, Orth, 2000). Table 2 lists the two sets of tasks, screen was used to deliver video prompting.
which were paired by difficulty level and motor response
requirements. The DVD video prompts were made using an adult
model and included verbal cues corresponding to each
The instructor conducted baseline, comparison, and fi- step of the task analysis. For example, the video showed
nal treatment sessions in the home living room of the the adult model turning the kitchen timer to 20 minutes
students’ local high school. The home living room con- and saying, “Turn the dial to 20.” During prompting ses-
tained a 20 ft. x 15 ft. kitchen area with a 39 in. x 95 in. sions, the instructor pushed the “Pause” button at the
table where all materials were placed. The stove and sink end of each video segment and the student completed
were located in front of the table. The student sat next to the task. The video prompts ranged from 12 s to 25 s.
the instructor at the table, and the reliability data collec-
tor sat at the end of the table. A portable DVD player was Static Picture Prompting. Identical to the Mechling and
positioned on the table to the left of the student, and the Gustafson (in press) study, laminated 4 in. x 7 in. single
instructor held up static picture cards as appropriate. photographs or line drawings were used to present the
static picture prompts to students and were constructed
based on those found in three commercially available
picture cookbooks (Look’n Cook Microwave: Easy-to-
Make Illustrated Recipes; Stepwise Lunch Cookbook; Visual
Recipes, A Cookbook for Non-Readers. Photo size approxi-
Table 2
mated the dimensions of the video image on the portable
DVD screen (7 in.).
Cooking-Related Tasks

Set 1 Tasks Set 2 Tasks Response Definitions and Data Collection

Coat chicken with season- Put chicken in the plastic
ing on plate. bag. Shake. The instructor collected data on each student’s ability
Set dial timer to 20 Turn oven dial to bake to perform each cooking-related task during each con-
minutes. 350°. dition. Data were reported based on the percentage of
Open crescent roll can. Pull tab to open frozen cooking-related tasks performed correctly by each stu-
orange juice. dent using pictures or video prompts. Response defi-
Sprinkle chicken with Sprinkle fish with lemon nitions were determined by observing adults without
paprika from can. juice (plastic lemon). disabilities completing the tasks. Responses were re-
Grate block of cheese with Peel carrot with vegetable corded as (a) correct – student initiated a task within 3
hand grater. peeler. s of presentation of the static picture or end of the video
Measure ¼ cup water from Measure ½ cup hot water prompt and performed the task independently correct
the sink. from coffee pot. within 1 min.; (b) incorrect–student failed to initiate
Use a whisk to stir milk. Scoop up dough with mel- the task within 3 s of presentation of the static picture
lon/cookie scoop. or end of the video prompt; (c) incorrect (topographical
Cut off each end of celery Snap off ends of asparagus. error)–student failed to complete the task correctly; (d)
with knife. incorrect (duration error)–student failed to complete the
Put paper liners in muffin Grease loaf pan with stick step within 1 min. of presentation of the static picture
pan. of butter. or end of the video prompt. Experimental control was
Crack two eggs open into Use fork to prick sweet determined through visual analysis of the data showing
bowl. potato. consistent and distinctly different levels of responding

34 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

across the two interventions (Kennedy, 2005) and com- conducted individually. During each condition, one ses-
parison of mean scores. sion was implemented each afternoon (two-three after-
noons per week). A set of 10 static picture prompting
tasks and a set of 10 video prompting tasks were present-
Experimental Design ed during each session (one trial per task). One instruc-
An adapted alternating-treatments design (AATD) with tor conducted all sessions, and sets were counterbalanced
baseline, comparison, withdrawal, and final treatment across sessions so that sets of static pictures and video
conditions was used to compare the effects of static pic- prompts were alternately presented during respective
ture and video prompting on the performance of cook- sessions.
ing-related tasks by six students with ASD (Richards,
Taylor, Ramasamy, & Richards, 1999). The two inter-
ventions were applied to different but equally difficult
Baseline Procedures
sets of tasks. Equivalency of tasks was based on the motor Students’ abilities to complete each of the 20 cooking-
response requirements of the tasks, which were paired related tasks were evaluated over four baseline sessions
across two sets of behaviors (see Table 2). or until data stabilized (Tawney & Gast, 1984) prior
to the first comparison session. Data were collected to
The two sets of behaviors (cooking-related tasks) were determine each student’s levels of performance on the
counterbalanced across interventions and two groups tasks prior to implementing the two interventions. Each
of students, as shown in Table 1. Set 1 was assigned to session consisted of one trial per cooking-related task.
three students using video prompting, whereas Set 2 was A verbal task direction such as “Use the fork to prick
assigned to the same three students using static picture the sweet potato” (see Table 3) was provided, and the
prompts. The reverse order of sets and interventions was student was given 3 s to initiate the task and 1 min. to
assigned to the second group of three students. Order complete it. In addition to the 20 task-related materials
of presentation of prompting systems was counterbal-
anced within sessions to reduce the impact of possible
confounding variables, carryover, and order effects. In-
structor, time of day, and setting remained constant. The Table 3
dependent variable was the percent of cooking-related
tasks completed independently. Verbal Directions for Cooking-Related Tasks

Set 1 Directions Set 2 Directions

Experimental conditions included baseline (verbal task
direction), static picture and video prompting (compari- “Coat chicken with “Put chicken in the plastic
son condition), follow-up, and a final treatment condi- seasoning.” bag. Shake.”
tion. The follow-up condition immediately followed the “Set timer to 20 minutes.” “Turn oven on to bake
comparison condition in order to evaluate students’ abil- 350°.”
ities to complete all 20 tasks without visual prompting. “Open rolls.” “Pull tab to open orange
The final treatment condition assessed possible multiple- juice.”
treatment interference (drop in performance during this “Sprinkle chicken with “Spinkle fish with lemon
phase may indicate interference) and to evaluate the ef- paprika.” juice.”
fects of the most effective treatment on all 20 tasks. “Grate the cheese.” “Peel carrot.”
“Measure ¼ cup water “Measure ½ cup hot water
from the sink.” from pot.”
General Procedures
“Use a whisk to stir milk.” “Scoop up dough.”
Students were assigned to one set for video prompting “Cut off each end of “Snap off ends of
and a comparable set for static picture prompting (see celery.” asparagus.”
Table 1). Cooking-related tasks were paired to replicate “Put paper liners in pan.” “Grease loaf pan.”
those of Mechling and Gustafson (in press), who paired “Get two eggs. “Crack “Use fork to prick the
tasks by level of motor response difficulty (see Table 2). open in the bowl.” sweet potatoes.”
Baseline, comparison, and final treatment sessions were

JSET 2008 Volume 23, Number 3 35

 Journal of Special Education Technology

placed on the table, 10 non-examples (to avoid process ing this condition. Instead, the sets previously receiving
of elimination as target items were used) were placed on static picture prompting were now presented with video
the table. Students received nonspecific verbal praise on prompts.
the average of every third task (VR3) for general attend-
ing, attempts to perform the tasks, and correct responses.
Students later cooked and consumed crescent rolls and
Reliability
juice when containers were correctly opened. A university student collected interobserver agreement
and procedural reliability data simultaneously on 20%
of all sessions across conditions. Interobserver agreement
Static Picture and Video Prompting was recorded on student responses to set one and set two
Comparison Procedures tasks (Table 2) and calculated using the point-by-point
Following baseline, prompting sessions began for each method and dividing the number of instructor and ob-
student. Each trial started with a verbal task direction server agreements by number of agreements plus dis-
identical to those used during baseline. The student was agreements and multiplying by 100. Mean interobserver
then shown a static picture of the task (e.g., scooping agreement when recording student responses was 99.2%
cookie dough) or a video model (e.g., squeezing lemon across all participants and conditions (range = 95-100).
on fish sticks) paired with verbal directions presented on
the video model. The student was given 3 s to initiate Procedural reliability data were collected on the follow-
each task and 1 min. to complete it. If the student failed ing instructor behaviors: (a) presenting the static picture
to initiate the task, failed to complete the task within and delivering the task direction; (b) turning on the
1 min. or incorrectly performed the task, the instructor DVD player and delivering the task direction; (c) having
presented the next static picture or video prompt. The all materials ready; (d) waiting 3 s for task initiation and
procedure continued until all 20 cooking-related tasks 1 min. for task completion; (e) no delivery of prompts or
were presented (one trial for each set of 10 video prompt- cues; and (f ) delivery of reinforcement. Procedural reli-
ing tasks and one trial for each set of 10 static picture ability agreement was calculated by dividing the number
prompting tasks). Ten trials were presented consecutively of observed instructor behaviors by the planned number
for static pictures or video prompts. Students received of behaviors and multiplying by 100 (Billingsley, White,
nonspecific verbal praise on the average of every third & Munson, 1980). Mean procedural agreement was
task (VR3) for general attending, attempts to perform 99.7% (range = 98.3-100). Procedural errors consisted
tasks, and correct responses. Comparison procedures of the following: crescent rolls opened prematurely, cres-
continued for a minimum of eight sessions or until data cents rolls did not open when student removed wrapper,
stabilized and an experimental effect was demonstrated a different bowl was placed on the table for the eggs, and
in favor of one procedure (Tawney & Gast, 1984). intercom school announcements interrupted a trial.

Follow-Up Procedures
Immediately after the comparison phase of the study, the Results
20 tasks were individually evaluated to determine wheth-
Figures 1 and 2 present the performance data compar-
er students could complete tasks with only verbal task
ing the effects of static picture and video prompting on
directions following static picture or video prompting.
the completion of cooking-related tasks by each of the
Procedures were identical to those used during baseline.
six students with ASD. Results are consistent with the
replicated study of Mechling and Gustafson (in press),
Final Treatment Procedures which indicated that video was more effective than static
pictures in prompting task completion by students with
During the final treatment condition, the single most ef-
disabilities.
fective intervention (video prompting) was applied to all
20 cooking-related tasks for four sessions or until data
Although the six students had previous experience with
stabilized. Procedures were identical to those used dur-
cooking-related tasks, only two were able to complete a
ing video prompting. No static pictures were used dur-
small percentage of tasks during baseline when only a

36 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Figure 1

Percent correct performance for Group 1 (Jessie, Kyle, Josh) across conditions and treatments
B as eline Com paris on W ithdrawl
V erbal Direc tion V ideo/S tatic P ic ture V erbal Direc tion F inal Treatm ent

100
V ideo

80 P ic ture
P e rc e nt C o rre c t

60

40

20
J e s s ie
0

100

80
P e rc e nt C o rre c t

60

40

20
K y le

100

80
P e rc e nt C o rre c t

60

40

20
Josh
0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

S e ssio ns

Note: Closed circles = static picture prompts; open squares = video prompts. Closed circles in final treatment:
tasks receiving static picture prompts in comparison phase received video prompts in final condition.

JSET 2008 Volume 23, Number 3 37

 Journal of Special Education Technology

Figure 2

Percent correct performance for Group 2 (Gary, Eddie, Dan) across conditions and treatments
B as eline Com paris on W ithdrawl
V erbal Direc tion V ideo/S tatic P ic ture V erbal Direc tion F inal Treatm ent

100
V ideo

80 P ic ture
P e rc e nt C o rre c t

60

40

20
G a ry
0

100

80
P e rc e nt C o rre c t

60

40

20
E ddie

100

80
P e rc e nt C o rre c t

60

40

20
Dan
0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

S e ssio ns
Note: Closed circles = static picture prompts; open squares = video prompts. Closed circles in final treatment:
tasks receiving static picture prompts in comparison phase received video prompts in final condition.

38 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

verbal task direction was provided, and four of the stu- performance with video prompting during the compari-
dents were unable to complete any of the tasks during son phase of the study.
baseline. Increases were noted for both static pictures
and video prompting immediately upon their introduc- Figure 3 contains comparison of means for each student
tion in the comparison condition, although performance per treatment. The increase and difference between the
with static pictures remained relatively low for Josh, Jes- two prompting procedures was most noticeable for Jessie
sie, and Dan. In comparison, video prompting resulted and Josh. Jessie’s mean correct performance was 85% for
in a rapid improvement in task completion for all six video prompting and 15% for static pictures, whereas
students with a consistent difference in the level of per- Josh’s mean correct performance was 86.3% for video
formance for all students in favor of video prompting. prompting and 18.8% for static pictures. A significant
Jessie was the only student who reached 100% correct difference also occurred between the two prompting
procedures for Kyle (mean performance: 76% video

Figure 3

Mean percentage scores for each student, group, and all students per treatment

JSET 2008 Volume 23, Number 3 39

 Journal of Special Education Technology

prompting, 35% static picture prompting) and Dan video prompting (mean 70%). Dan’s mean increase was
(mean performance: 57.5% video prompting, 20% static 42.5% (static picture comparison–20%; video prompt-
picture prompting). ing applied in final treatment–62.5%) while Kyle’s mean
increase was 30% (static picture comparison–35%; video
The mean performance was greater for each group of stu- prompting applied in final treatment–65%). Gary’s per-
dents when using video prompting, regardless of the set formance with the original static picture set increased
(see Figure 3). The mean for Group 1 was 74.4% for vid- above his performance with his video prompting set of
eo prompting and 55.2% for static picture prompting, tasks during the final treatment condition. He main-
whereas the mean scores for Group 2 were 81.9% for tained a mean performance of 80% correct with the
video prompting and 23.9% for static picture prompt- video prompting set of tasks and increased his mean per-
ing. Overall, video prompting resulted in a higher formance to 82.5% when video prompting was applied
percentage of correct task performance across the six stu- to the static picture set.
dents (78.2% mean correct) compared to static pictures
(39.22% mean correct) during the comparison phase of When video prompting was applied to static picture sets
the study. for the other five students, their performance remained
lower than for original video prompting sets. The differ-
ence in performance was likely due to the design of the
Follow-Up study, which included only four sessions in the final treat-
During the follow-up condition, four of the six stu- ment phase, whereas video prompting was applied for a
dents maintained performance levels when static picture minimum of eight sessions during the comparison phase.
prompts were removed, but Eddie and Kyle demon-
strated a slight drop in performance. In comparison,
three of the six students’ (Jessie, Josh, and Dan) per-
formance dropped significantly when video prompting Discussion
was removed. Indeed, Dan refused to complete his set of
The current study extends the limited literature compar-
tasks when the DVD player and video prompting were
ing the effects of video and static picture prompts. To
removed.
date, a small number of studies have compared the ef-
fectiveness of using static pictures and video technology
Final Treatment to teach students with disabilities (Alberto et al., 2005;
Cihak et al., 2006; Mechling & Gustafson, in press; Ste-
In the final treatment phase, only the most effective treat-
phens & Ludy, 1975). Results support the findings of
ment (video prompting) was administered. All students
Mechling and Gustafson, who found that students with
maintained their level of performance with sets of tasks
moderate intellectual disabilities independently com-
receiving video prompting in the comparison phase,
pleted a greater percentage of tasks when using video
with Eddie increasing his level of performance to 100%
prompting compared to static picture prompts, and ex-
in two of four sessions during final treatment. Jessie also
tend the results to students with ASD.
completed 100% of the tasks correctly in three of four
sessions during final treatment.
The results of the current study further support research
suggesting that there is a high degree of similarity be-
When video prompting was applied to tasks that received
tween videotape and real-life experiences (Alcantara,
static picture prompting in the comparison phase, all six
1994) and that video technology provides a realistic way
students demonstrated an increase in performance com-
to closely replicate stimulus and response requirements
pared to their performance using static pictures. Jessie
for task completion. Video technology provides visual
demonstrated the greatest increase in performance from
supports that combine audio and video recordings that
static picture prompting (mean correct 15%) when video
can be presented repetitively and provide predictability
prompting was applied to the same tasks (mean correct
to learners. Unlike adult delivered instruction, which
87.5%). Josh demonstrated a 51.3% mean increase in
cannot control all variables (e.g., interruptions, variabil-
performance from static pictures in the comparison phase
ity of substitute teachers), video technology can provide
(mean 18.8%) to performance in final treatment with

40 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

instruction that is identical each time and that can be a video may positively affect the performance of students
played over and over (Neumann, 2004). with ASD.

Video technology may further allow focus on salient The current study supported the use of video compared to
features of instruction without the distraction of human static pictures to prompt task completion, yet one system
interaction (sounds, body language, smells) for students does not fit all students; therefore, future research should
with ASD, who may have difficulty focusing or may be further examine student characteristics as they relate to
uncomfortable with human interactions that tend to be the effectiveness of one procedure compared to another.
emotionally laden (Neumann, 2004; Stephens & Ludy, Students with ASD are characterized as being stronger
1975). Video allows for filtering of extraneous variables visual learners (Quill, 1997), which supports the use of
and students with ASD can focus on a small spatial area both picture and video systems. The results of a com-
of the screen and minimize attentional and language re- parison of baseline performance and performance with
quirements (Sherer et al., 2001). By zooming in on sa- static pictures and video modeling in the current study
lient features important to the task, video recordings can indicated that instructional procedures that incorporated
help compensate for students’ tendency toward stimulus visual information were more effective than verbal cues
over-selectivity when they are no longer able to view a alone for persons with ASD (Quill, 1997; Vaughn &
multitude of images (Charlop-Christy, et al., 2000). Horner, 1995). Tiong, Blampied, and Le Grice (1992)
suggest that one advantage of video over picture prompt-
Although the current study was supportive of video ing lies in the ability of video to present information in
technology for prompting task completion, limitations real time, dynamic order. They also note that movement
do exist. All students did not perform at 100% accuracy in video makes cues more salient and directs attention to
during the comparison or final treatment phases. Perfor- key and natural stimuli. However, in order to use video
mance during the final treatment was likely due to the systems, students need to be imitative, able to attend to
limitation of four sessions. Performance below 100% the model, and able to understand what is occurring on
correct may also suggest that the individual learning the screen (Meharg & Woltersdof, 1990). A question re-
styles of students with ASD must be considered. Video mains concerning the level of imitative ability students
prompting requires some shifting of attention due to the must possess. Will students with emerging imitation,
use of motion. Verbal instructions were also included in who still require hand-over-hand prompting benefit
the video prompting condition. Each may have affected from video systems? Can video modeling be paired with
student performance. physical prompting for students who do not possess a
high level of imitative skills?
Research suggests that the language requirements and
speed of delivery of information in commercially available A further limitation of the study was task length. Each
computer and television programs may be too advanced task required a minimal number of steps rather than
for students with ASD (Neumann, 2004). The current multiple steps typically required for cooking or comple-
video programs were designed to prompt all six students tion of functional tasks. Types of skills and response re-
using the same programs. Future research should explore quirements may influence results. Video instruction can
how to create instructional video programs that can be provide multiple and varying examples required in teach-
catered to the learning style of the student. By attending ing social role plays (Simpson, Langone, & Ayres 2004)
to the specific needs of the learner, parents or teachers and multiple examples of verbal and motor responses
can develp videotaped instruction or prompting pack- required in complex communication skills (D’Ateno,
ages that slow the pace, filter out distracting, extrane- et al., 2003). However, picture prompts may provide a
ous information, and simplify the language (Neumann, portable system for tasks involving a complex set of skills
2004). Gepner, Deruelle, & Grynfeltt (2001) suggest that need to be completed in a specific sequence. Future
that the speed of motion rather than the motion per se research should compare the effects of static picture and
may influence student performance, favoring slower mo- video prompting strategies when teaching multiple and
tion. Further, Nikopoulos and Keenan (1993) also pro- complete task activities such as those of daily living or vo-
pose that a reduced number of behavioral components in cational tasks. This is especially important given the lim-
ited research comparing static pictures and video and the

JSET 2008 Volume 23, Number 3 41

 Journal of Special Education Technology

contrasting results of Alberto, et al. (2005), and Cihak, Future research should continue to examine the com-
et al. (2006). Additionally, based on three commercially bined effects of both video and pictures. While Van Laar-
available picture cook books, only one photograph was hoven and Van Laarhoven-Meyers (2006) found video
used to depict a task. Multiple pictures representing mul- and video plus photos to be effective in teaching daily
tiple steps within a task could influence task completion living skills, video plus photos was more efficient in terms
and should, therefore, be evaluated in future studies. of number of required sessions to reach criterion than
video alone. Computer-based programs have combined
Questions remain not only concerning whether stu- pictures and video in an interactive learning environment
dents with ASD learn better with one procedure than to teach multi-step tasks (Hagiwara & Myles, 1999;
the other, but whether they have a preference for using Mechling & Gast, 2003; Mechling, Gast, & Barthold,
pictures or video. Because lack of motivation can impede 2003) and include pictures, video, text captioning, and
learning new skills (Charlop-Christy & Haymes, 1996), interactive features that make learning “come alive” for
it is important to identify the reinforcing properties of persons with disabilities (Williams & McClintic, 2005).
instructional materials, which can be used to teach new As students’ independence increases, a combined system
skills. Perhaps the exceptional memory and intrinsic mo- may be a better match for some task requirements. Video
tivation for videos, television shows, commercials, and may be faded while a permanent picture system remains.
movies (Charlop-Christy, Le, & Freeman, 2000; Neu- In the current study, pictures remained on the table while
mann, 2004; Sherer et al., 2001), especially when the students completed tasks, but the portable DVD player
content is something in which they are interested, may showed a blank screen at the end of each video prompt-
compensate for any weaknesses in imitative skills if stu- ing segment. While features such as a still frame could be
dents are highly motivated to repeatedly watch the con- incorporated into future programs, the frame at the end
tent of video instruction. of a video segment may not provide prompts equivalent
to a photograph that has been carefully created to depict
Motivation to watch video images may also have im- as much information for the learner as possible.
plications for creating instructional videos. Personally
created video programs can imbed the interest of a stu- In summary, both picture based and video prompts have
dent (e.g., power tools) as a reinforcement at the end been demonstrated to be effective in teaching skills to
of an instructional video or within the content of the students with ASD and prompting task completion.
video to reinforce task completion (Mechling, Gast, & There is no single best method to teach students with
Cronin, 2006). Although studies have assessed and iden- ASD (Simpson 2005). However, teachers, parents, and
tified stimuli for use as reinforcement for students with practitioners have expressed a need for effective practic-
ASD (Carr, Nicolson, & Higbee, 2000; Mason, McGee, es that are scientifically based (Simpson) and that ade-
Farmer-Dougan, & Risley, 1989), preference of stimuli quately assess the relationship between independent and
for delivering instruction requires further investigation. dependent variables (Wheeler, Baggett, Fox, & Blevins,
2006). While this study added to the growing body of
A form of video technology, video self-modeling (VSM), literature and interest in comparing the effects of static
may possess particular reinforcing value for persons pictures and video modeling, the findings suggest ad-
with ASD. Video self-models present video segments ditional avenues of study to maximize the potential of
of the student positively performing a task at advanced technology.
levels (Dowrick, 1999), which may increase feelings of
self-esteem and beliefs that he or she can be successful
(Buggey, 2005). Students with ASD may be more moti-
vated by watching themselves, and the familiarity of self References
may make learning and processing of visual information Agran, M., Fodor-Davis, J., Moore, S., & Deer, M. (1989). The ap-
easier (Sherer et al., 2001). In their review of videotape plication of a self-management program on instruction-following
self-modeling, Meharg and Woltersdorf (1990) report skills. Journal of the Association for the Severely Handicapped, 14,
findings that subjects were more likely to imitate be- 147-154.
haviors when they perceived the model to be similar to
themselves.

42 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Alberto, P., Cihak, D. F., & Gama, R. I. (2005). Use of static picture Cihak, D., Alberto, P. A., Taber-Doughty, T., & Gama, R. I. (2006).
prompts versus video modeling during simulation instruction. Re- A comparison of static picture prompting and video prompting
search in Developmental Disabilities, 26, 327-339. simulation strategies using group instructional procedures. Focus
on Autism and Other Developmental Disabilities, 21, 89-99.
Alcantara, P. R. (1994). Effects of videotape instructional package on
the purchasing skills of children with autism. Exceptional Chil- Dauphin, M., Kinney, E. M., & Stromer, R. (2004). Using video
dren, 61, 40-55. enhanced activity schedules and matrix training to teach socio-
dramatic play to a child with autism. Journal of Positive Behavior
Apple, A. L., Billingsley, F., & Schwartz, I. S. (2005). Effects of video
Interventions, 6, 238-250.
modeling alone and with self-management on compliment-giving
behaviors of children with high functioning autism. Journal of D’Ateno, P., Mangiapanello, K., & Taylor, B. A. (2003). Using video
Positive Behavior Interventions, 7, 33-46. modeling to teach complex play sequences to a preschooler with
autism. Journal of Positive Behavior Interventions, 5, 5-12.
Ayres, K. M., & Langone, J. (2005). Intervention and instruction
with video for students with autism: A review of the literature. Ed- Delano, M. E. (2007). Video modeling interventions for individuals
ucation and Training in Developmental Disabilities, 40, 183-196. with autism. Remedial and Special Education, 28, 33-42.
Bellini, S., & Akullian, J. (2007). A meta-analysis of video modeling Dettmer, S., Simpson, R. L., Myles, B. S., & Ganz, J. B. (2000).
and video self-modeling interventions for children and adoles- The use of visual supports to facilitate transitions of students with
cents with autism spectrum disorders. Exceptional Children, 73, autism. Focus on Autism and Other Developmental Disabilities, 15,
264-287. 163-169.
Billingsley, F. F., White, O. R., & Munson, R. (1980). Procedural Dowrick, P. W. (1999). A review of self modeling and related inter-
reliability: A rationale and an example. Behavioral Assessment, 2, ventions. Applied & Preventative Psychology, 8, 23-39.
229-241.
Gepner, B. Deruelle, C., & Grynfeltt, S. (2001). Motion and emo-
Bryan, L. C., & Gast, D. L. (2000). Teaching on-task and on-sched- tion: A novel approach to the study of face processing by young
ule behaviors to high-functioning children with autism via picture autistic children. Journal of Autism and Developmental Disorders,
activity schedules. Journal of Autism and Developmental Disorders, 31, 37-45.
30, 553-567.
Hagiwara, T., & Myles, B. S. (1999). A multimedia social story inter-
Buggey, T. (2005). Video self-modeling applications with students vention: Teaching skills to children with autism. Focus on Autism
with autism spectrum disorder in a small private school setting. and Other Developmental Disabilities, 14, 82-95.
Focus on Autism and Other Developmental Disabilities, 20, 52-63.
Hall, L. J., McClannahan, L. E., & Krantz, P. J. (1995). Promoting
Buggey, T., Toombs, K., Gardener, P., & Cervetti, M. (1999). Train- independence in integrated classrooms by teaching aides to use ac-
ing responding behaviors in students with autism: Using video- tivity schedules and decreased prompts. Education and Training in
taped self-modeling. Journal of Positive Behavior Interventions, 1, Mental Retardation and Developmental Disabilities, 30, 208-217.
205-214.
Hanson, D. (1999). Look’n cook microwave: Easy-to-make illustrated
Carr, J. E., Nicolson, A. C., & Higbee, T. S. (2000). Evaluation of recipes. Verona, WI: Attainment Company.
brief multiple-stimulus preference assessment in a naturalistic
Haring, T. G., Kennedy, C. H., Adams, M. J., Pitts-Conway, V.
context. Journal of Applied BehaviorAnalysis, 33, 353-357.
(1987). Teaching generalization of purchasing skills across com-
Charlop-Christy, M. H., Carpenter, M., Le, L., LeBlanc, L. A., & munity settings to autistic youth using videotape modeling. Jour-
Kellet, K. (2002). Using the picture exchange communication nal of Applied Behavior Analysis, 20, 89-96.
system (PECS) with children with autism: Assessment of PECS
Higgins, K., & Boone, R. (1996). Creating individualized computer-
acquisition, speech, social-communicative behavior, and problem
assisted instruction for students with autism using multimedia
behavior. Journal of Applied Behavior Analysis, 35, 213-231.
authoring software. Focus on Autism and Other Developmental Dis-
Charlop-Christy, M. H., & Daneshvar, S. (2003). Using video mod- abilities, 11, 69-78.
eling to teach perspective taking to children with autism. Journal
Jackson, B. (1998). Stepwise lunch cookbook. Framingham, MA: Ther-
of Positive Behavior Interventions, 5, 12-10.
apro, Inc.
Charlop-Christy, M. H., & Haymes, L. K. (1996). Using obsessions
Kennedy, C. H. (2005). Single-case designs for educational research.
as reinforcers with and without mild reductive procedures to de-
Boston: Pearson, Allyn and Bacon.
crease inappropriate behaviors of children with autism. Journal of
Autism and Developmental Disorders, 26, 527-546. Kinney, E. M., Vedora, J., & Stromer, R. (2003). Computer-present-
ed video models to teach generative spelling to a child with an au-
Charlop-Christy, M. H., Le, L., & Freeman, K. A. (2000). A com-
tism spectrum disorder. Journal of Positive Behavior Interventions,
parison of video modeling with in vivo modeling for teaching
5, 22-29.
children with autism. Journal of Autism and Developmental Dis-
orders, 30, 537-552. Krantz, P. J., MacDuff, M. T., & McClannahan, L. E. (1993). Pro-
gramming participation in family activities for children with au-
Charlop, M. H., & Milstein, J. P. (1989). Teaching autistic children
tism: Parents’ use of photographic activity schedules. Journal of
conversational speech using video modeling. Journal of Applied
Applied Behavior Analysis, 26, 137-138.
Behavior Analysis, 22, 275-285.

JSET 2008 Volume 23, Number 3 43

 Journal of Special Education Technology

Lancioni, G. E., & O’Reilly, M. F. (2001). Self-management of in- Pierce, K. L., & Schreibman, L. (1994). Teaching daily living skills
struction cues for occupation: Review of studies with people with to children with autism in unsupervised settings through picto-
severe and profound developmental disabilities. Research in Devel- rial self-management. Journal of Applied Behavior Analysis, 27,
opmental Disabilities, 22, 41-65. 471-481.
MacDuff, G. S., Krantz, P. J., & McClannahan, L. E. (1993). Teach- Quill, K. A. (1997). Instructional considerations for young children
ing children with autism to use photographic activity schedules: with autism: The rationale for visually cued instruction. Journal of
Maintenance and generalization of complex response chains. Jour- Autism and Developmental Disorders, 27, 697-714.
nal of Applied Behavior Analysis, 26, 89-97.
Reagon, K. A., Higbee, T. S., & Endicott, K. (2006). Teaching pre-
Mason, S. A., McGee, G. G., Farmer-Dougan, V., & Risley, T. R. tend play skills to a student with autism using video modeling
(1989). A practical strategy for ongoing reinforcer assessment. with a sibling as model and play partner. Education and Treatment
Journal of Applied Behavior Analysis, 22, 171-179. of Children, 29, 517-528.
Massey, N. G., & Wheeler, J. J. (2000). Acquisition and generaliza- Richards, S. B., Taylor, R. L., Ramasamy, R., & Richards, R. Y.
tion of activity schedules and their effects of task engagement in (1999). Single subject research: Applications in educational and clin-
a young child with autism in an inclusive pre-school classroom. ical settings. San Diego, CA: Singular Publishing Group, Inc.
Education and training in Mental Retardation and Developmental
Schmit, J., Alper, S., Raschke, D., & Ryndak, D. (2000). Effects
Disabilities, 35, 326-335.
of using a photographic cueing package during routine school
Mechling, L. C. (2007). Assistive technology as a self-management transitions with a child who has autism. Mental Retardation, 38,
tool for prompting students with intellectual disabilities to initiate 131-137.
and complete daily tasks: A literature review. Education and Train-
Schwartz, I. S., Garfinkle, A. N., & Bauer, J. (1998). The picture
ing in Developmental Disabilities, 42, 252-269.
exchange communication system: Communicative outcomes for
Mechling, L. C. (2005). The effect of instructor-created video pro- young children with disabilities. Topics in Early Childhood Special
grams to teach students with disabilities: A literature review. Jour- Education, 18, 144-159.
nal of Special Education Technology, 20(2), 25-36.
Sherer M., Pierce, K. L., Paredes, S., Kisacky, K. L., Ingersoll, B.,
Mechling, L. C., & Gast, D. L. (2003). Multi-media instruction to & Schreibman, L. (2001). Enhancing conversation skills in chil-
teach grocery word associations and store location: A study of dren with autism via video technology. Behavior Modification, 25,
generalization. Education and Training in Mental Retardation and 140-158.
Developmental Disabilities, 38, 62-76.
Shipley-Benamou, R., Lutzker, J. R., & Taubman, M. (2002). Teach-
Mechling, L. C., Gast, D. L., & Barthold, S. (2003). Multi-media ing daily living skills to children with autism through instruc-
computer-based instruction to teach students with moderate in- tional video modeling. Journal of Positive Behavior Interventions,
tellectual disabilities to use a debit card to make purchases. Excep- 4, 165-175.
tionality, 11, 239-254.
Simpson, R. L. (2005). Evidence-based practices and students with
Mechling, L. C., Gast, D. L., & Cronin, B. A. (2006). The effects autism spectrum disorders. Focus on Autism and Other Develop-
of presenting high-preference items, paired with choice, via mental Disabilities, 20, 140-149.
computer-based video programming on task completion of stu-
Simpson, A., Langone, J., & Ayres, K. (2004). Embedded video and
dents with autism. Focus on Autism and Other Developmental Dis-
computer based instruction to improve social skills for students
abilities, 21, 7-13.
with autism. Education and Training in Mental Retardation and
Mechling, L. C., & Gustafson, M. (in press). Comparison of the Developmental Disabilities, 39, 240-252.
effects of static picture and video prompting on completion of
Stephens, W. E., & Ludy, I. E. (1975). Action-concept learning in
cooking related tasks by students with moderate intellectual dis-
retarded children using photographic slides, motion picture se-
abilities. Exceptionality.
quences, and love demonstrations. American Journal of Mental
Meharg, S. S., & Woltersdorf, M. A. (1990). Therapeutic use of vid- Deficiency, 80, 277-280.
eotape self-modeling: A review. Advances in Behaviour Research
Stromer, R., Kimball, J. W., Kinney, E. M., & Taylor, B. A. (2006).
and Therapy, 12, 85-99.
Activity schedules, computer technology, and teaching children
Neumann, L. (2004). Video modeling: A visual teaching method for with autism spectrum disorders. Focus on Autism and Other Devel-
children with autism. Brandon, FL: Willerik Publishing. opmental Disabilities, 21, 14-24.
Nikopoulos, C. K., & Keenan, M. (2003). Promoting social initia- Sturmey, P. (2003). Video technology and persons with autism and
tion in children with autism using video modeling. Behavioral In- other developmental disabilities. Journal of Positive Behavior Inter-
terventions, 18, 87-108. ventions, 5, 3-4.
Nikopoulos, C. K., & Keenan, M. (2004). Effects of video modeling Tawney, J. W., & Gast, D. L. (1984). Single subject research in special
on social initiations by children with autism. Journal of Applied education. Columbus, OH: Merrill.
Behavior Analysis, 37, 93-96.
Taylor, B. A., Levin, L., & Jasper, S. (1999). Increasing play-related
Orth, T. (2000). Visual recipes: A cookbook for non-readers. New York: statements in children with autism toward their siblings: Effects of
DRL Books, Inc. video modeling. Journal of Developmental and Physical Disabilities,
11, 253-264.

44 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Thiemann, K. S., & Goldstein, H. (2001). Social stories, written cues, Williams, J. M., & McClintic, H. (2005). Digital technology offers
and video feedback: Effects on social communication of children new opportunities to people with disabilities. The Exceptional Par-
with autism. Journal of Applied Behavior Analysis, 34, 425-446. ent, 35, 34-41.
Tincani, M. (2004). Comparing the picture exchange communication Wilson, P. G., Schepis, M. M., & Mason-Main, M. (1987). In vivo
system and sign language training for children with autism. Focus use of picture prompt training to increase independent work at
on Autism and Other Developmental Disabilities, 19, 152-163. a restaurant. Journal of the Association for Persons with Severe Dis-
abilities, 12, 145-150.
Tiong, S. J., Blampied, N. M., & Le Grice, B. L. (1992). Training
community-living, intellectually handicapped people in fire safety
using video prompting. Behavior Change, 9, 65-72. Author Notes
Van Laarhoven, T., & Van Laarhoven-Myers, T. (2006). Comparison
of three video-based instructional procedures for teaching daily Linda C. Mechling is an associate professor of special education,
living skills to persons with developmental disabilities. Education University of North Carolina–Wilmington. Melissa R.Gustafson
and Training in Developmental Disabilities, 41, 35-381. is a graduate student seeking a teaching license in the General
Curriculum Special Education Program, University of North
Vaughn, B., & Horner, R. H. (1995). Effects of concrete versus verbal
choice systems on problem behavior. Augmentative and Alternative Carolina–Wilmington.
Communication, 11, 89-92. Correspondence should be addressed to Linda C. Mechling, De-
Wert, Y. B., & Neisworth, J. T. (2003). Effects of video self-modeling partment of Early Childhood and Special Education, University
on spontaneous requesting in children with autism. Journal of of North Carolina–Wilmington, 601 S. College St., Wilmington,
Positive Behavior Interventions, 5, 30-35. NC 28403-5940. Email to [email protected]
Wheeler, J. J., Baggett, B. A., Fox, J., & Blevins, L. (2006). Treat-
ment integrity: A review of intervention studies conducted with
children with autism. Focus on Autism and Other Developmental
Disabilities, 21, 45-54.

JSET 2008 Volume 23, Number 3 45

 Journal of Special Education Technology

46 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Using Video Modeling to Increase Variation in the

Conversation of Children with Autism
Marjorie H. Charlop
Laura Gilmore
Claremont McKenna College

Gina T. Chang
Claremont Graduate University

The present study assessed the effects of video modeling on acquisition and generalization of variation in
the conversational speech of two boys with autism. A video was made showing several versions of several
topics of conversation, thus providing multiple exemplars of each conversation. Video modeling consisted
of showing each child a video depicting the various versions of the target conversation. A multiple baseline
design across children and within child was used. Generalization across setting, peer with autism, typically
developing peer, sibling at home, and topics of conversation were assessed. Ancillary measures of question-
asking and other social behaviors were taken in free play sessions. The results indicated that both subjects
increased variation in their conversation as a result of video modeling, as measured both within conversation
and cumulatively across conversations.

V ideo modeling has been investigated for more than

two decades as an effective way to teach a variety
of behaviors to persons with autism spectrum disorder
nally, maintaining a verbal exchange (Charlop-Christy
& Kelso, 2003). Although some of the video modeling
studies have examined aspects of conversational speech,
(ASD). This includes play skills, functional living skills, few have targeted the entirety of the behavioral chain re-
social behaviors, academic skills, and conversational skills quired for conversation.
(Apple, Billingsley, & Schwartz, 2005; Charlop-Christy,
Le, & Freeman, 2000; Charlop & Milstein, 1989; Niko- Rather, perhaps because of its complexity, conversational
poulos & Keenan, 2003; 2004). Few video modeling speech for persons with ASD has primarily been targeted
studies to date have addressed social communication through the use of scripts and cue cards (Charlop & Mil-
skills (Bellini & Akullian, 2007). Moreover, only two stein, 1989; Charlop-Christy & Kelso, 2003; Goldstein
studies have targeted the social communication skills of & Cisar, 1992; Krantz & McClannahan, 1993), which
conversational speech (Charlop & Milstein, 1989; Sherer allow the chain of responses necessary for a conversation
et al., 2001). to be broken down into discrete lines, thereby simplify-
ing it.
Unlike other skills within the social communication
domain (i.e., social initiations, verbal responses, and In one such study, Krantz and McClannahan (1993)
compliments), conversational speech requires the use implemented written scripts to facilitate social initia-
of multiple complex language skills such as initiation tions in four children with autism. During a structured
and expansion of a conversational topic, establishing an art project, participants were presented with scripts and
interactive “to-and-fro” pattern of conversation, and fi- were instructed to read initiation phrases to typically de-

JSET 2008 Volume 23, Number 3 47

 Journal of Special Education Technology

veloping peers in their proximity. The scripts were then ily teach the nonverbal aspects of a conversational inter-
systematically faded from end to beginning, until the action such as eye contact and body positioning.
scripts were no longer present. Results showed that when
the scripts were introduced, peer initiations increased, As mentioned, video modeling has been documented in
and as the scripts were faded, unscripted initiations in- the literature as being an effective technology to teach a
creased. While this research demonstrates the efficacy of variety of skills to individuals with ASD. It seems feasible
using written scripts to teach social initiations to persons that video modeling would be appropriate to teach con-
with ASD, initiations are only one part of the behav- versational speech as well (Charlop & Milstein, 1989).
ioral chain required to engage in conversational speech. Video modeling addresses the limitations of scripts and
Further, this procedure is limited to individuals who can cues by eliminating the requirements of reading abil-
read and carry out corresponding comprehension. ity, functioning level, and age. Moreover, video model-
ing is associated with additional advantages such as cost
Charlop-Christy and Kelso (2003) further investigated efficiency (Charlop & Milstein, 1989), rapid learning
the effectiveness of a cue card/written script program in (Charlop-Christy, Le, & Freeman, 2000), promotion
specifically teaching conversational speech to children of generalization (Matson & Sweezy, 1994), and abil-
with autism. Conversations were scripted and then bro- ity to present an entire conversation without the need
ken up into lines that were written on cue cards. The of fading.
conversant would say her line of the conversation and
then present the cue card for the child with autism to Two studies have examined the use of video modeling
read and respond. The conversant would subsequently to teach persons with ASD conversational skills (Char-
reply with her next line of conversation and present the lop & Milstein, 1989; Sherer et al., 2001). Sherer and
next cue card to the child to read and respond with, and colleagues (2001) investigated whether children with
so on. The presentation of the cue cards was gradually autism could learn a basic conversational exchange via
faded out until the conversant and the child engaged in video modeling. In this study, a conversational exchange
the short conversation without assistance. Generalization consisted of an answer response plus posing a question.
without the cue cards across people, setting, and an un- For example, the conversant may ask, “What is your fa-
trained topic occurred. Importantly, for two of the three vorite food,” whereas the modeled appropriate response
participants, response generalization occurred, such that for the child was “ My favorite food is pizza. What is your
after one conversation was taught using the cue cards, the favorite food?” Three out of five participants rapidly met
children began engaging in appropriate conversational criterion using video modeling, but two participants did
responses across other conversations (untrained scripts). not learn the behavior. All the children who met criteri-
Once again, however, this procedure may only be used on generalized the behavior across persons, settings, and
with children who can read and comprehend what they time. It is important to note, albeit not unexpected, that
are reading. these children were not able to generate new, untrained
responses to alternate generalization questions (i.e., re-
Although these studies show the potential of cue cards sponse generalization). That is, they could not vary their
and written scripts for teaching conversational skills, the responses to other topics of conversation.
applicability of these procedures is limited. For example,
as mentioned, these teaching methods require partici- In the second study, Charlop and Milstein (1989) taught
pants to be able to read and comprehend. The limitation conversational speech to three children with autism us-
of literacy also may reduce the age range of children who ing video modeling. Here, conversational speech was de-
can benefit from these techniques. Furthermore, while fined as the “to-and-fro” verbal exchange between two
the assumption that children with ASD would be on the people sustaining the interaction. The conversations that
higher functioning end of the spectrum because they were scripted and later modeled via video each consist-
would have the prerequisite verbal skills, many children ed of seven lines: four for the conversant and three for
with autism are able to engage in verbal exchanges but the child. The conversant typically started or ended the
are not able to read. Another potential limitation is that conversation. Each line in the conversation for the child
the use of cue cards and written scripts does not necessar- consisted of an answer and a question. This would sus-
tain the conversations. For example, the conversant in the

48 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

video asked, “What do you have?” The model replied, “A Method

box. Are you holding something?” The conversant then
said, “Yes, a box. What’s in your box?” The model re- Participants
sponded, “A ball. Is there something in your box?” and
so on. Thus, the study required the child to maintain Participants were two boys with autism who attended an
the topic of conversation and engage in multiple verbal after-school behavioral treatment program. Both chil-
exchanges. dren had been diagnosed with autism spectrum disorder
according to the Diagnostic and Statistical Handbook of
All three participants met criterion on the video mod- Mental Disorders, IV-TR, #299.00 (American Psychiat-
eling conversations, and all participants demonstrated ric Association [APA], 2000) by licensed psychologists
generalization across people, setting, and stimuli. For all at two independent agencies. Both were verbal and gen-
participants, exposure to the video modeling treatment erally answered questions but only occasionally asked
resulted in appropriate conversational speech across two questions. When asking questions, they typically asked
to four novel conversations. However, only one partici- the same questions repeatedly, which rarely served to
pant demonstrated generalization to an abstract conversa- maintain conversations. Both boys seldom generalized
tion (response generalization); the other two participants newly acquired speech skills, or other treatment gains, to
demonstrated partial use of conversational speech to an non-teaching settings. The two boys were considered to
abstract conversation. be on the mid to higher end of the autism spectrum due
to their speech skills.
In addition, the study assessed whether video modeling
would increase the children’s use of appropriate novel re- Gary was a 9-year-old boy, whose question-asking be-
sponses that varied from the modeled conversation. These havior was primarily rote and repetitive speech. He
ancillary data are important because video modeling has tended to ask the same few questions over and over,
been criticized for resulting in participants just learning such as “Where are we gonna go?” or “What are we
a rote way of responding. All participants increased their gonna do?” He rarely initiated conversations or spoke
response variation during the video modeling treatment, spontaneously. The majority of his speech was echolalic
and the variation continued to increase during general- or about topics he was preoccupied with, such as Honda
ization probes. These data are preliminary, however, and cars, nurses, and other unusual topics for a 9-year-old.
additional research is needed to address the use of video He engaged in stereotypy of finger rubbing, rocking,
modeling as a method to not only teach conversational and eye squinting.
speech but to simultaneously facilitate response variation
within conversational speech. Connor was an 8-year-old-boy who also had difficulty
maintaining conversations and seldom asked questions.
The present study was designed to expand previous re- He often engaged in inappropriate speech by persever-
search on using video modeling as a technology to teach ating on topics with which he was preoccupied, such as
conversational speech to children with autism; specifical- clouds and deserts. When asked a question, he generally
ly, to facilitate variation in conversational speech skills. gave a one- or two-word answer, although his speech
The study had three goals: (a) to systematically assess the about his topics of obsession indicated that he was ca-
effects of video modeling on increasing response varia- pable of much longer utterances.
tion in conversational speech; (b) to assess the effects of
video modeling on generalization of these varied conver- Conversants
sational skills across persons, settings, and topics of con-
versations; and (c) to measure any concomitant changes During baseline, treatment, testing, and probe conditions
in the children’s social behaviors and their amount of the conversant was the second author. During the gen-
question-asking in unstructured free play sessions. eralizations probes, the conversant was an experimenter,
a participant’s sibling, a therapist from the after-school
program, or a developmentally typical peer.

JSET 2008 Volume 23, Number 3 49

 Journal of Special Education Technology

Materials A5, so that while any one of these was about the robot
and typewriter, they all consisted of different lines so
Six sets of toys (e.g., Big Bird and Bert; a monster and a each child could potentially be presented with a num-
typewriter; a robot and a coloring book; a car and a jet ber of varying conversation scripts modeled and ide-
transformer; blue, red, and green yo-yos; and a black and ally generate varied responses. Each conversation topic
a red yo-yo) were chosen as stimuli to elicit conversa- also included two additional scripted conversations: (a)
tion. These toys were determined to be novel (no pre- Conversation A-test, which was designed to assess if the
vious contact) to the children based on parent, teacher, child learned conversational speech through video mod-
and therapist reports. Additionally, these toys were deter- eling and how much of that speech consisted of varied
mined to be of interest to the children based on pre-ex- responses; and (b) Conversation A', which was identical
periment choice assessments. That is, when individually to Conversation A, except that the toys that served as the
presented with other toys, the children chose these toys topic of conversation were changed so that the children
on two out three trials per toy. A variety of conversation had to make modifications in their responses to reflect
scripts about the toys were created. the ability to speak about the toys that were immediately
present and relevant, as opposed to merely memorizing
Conversation scripts. Each toy set served as the basis of a the lines from the scripted conversations that were mod-
conversation topic (see Table 1). For each conversation eled for them. For example, as seen in Table 1, Gary’s
topic, seven conversational scripts were written. As de- Conversation A is about yo-yos. His Conversation A' is
picted in Table 2, for each conversation topic (e.g., ro- about another pair of yo-yos that differed from the origi-
bot and typewriter) a set of five scripts was written and nal pair in color. Thus, Gary had to change some of his
identified by a letter and a number. The letter denoted conversation responses so that they were appropriate to
the topic of the conversation (e.g., Script A, robot and this new stimulus set of yo-yos. In other words, he had
typewriter) and the number denoted a particular ver- to talk about new features of these new yo-yos, and these
sion or variation of the script (A1, A2, A3, A4, A5). For conversation lines had not been modeled for him. This is
example, for Conversation A, about the robot and the an assessment of generalization across stimuli.
typewriter, there were five versions, A1, A2, A3, A4 and
Each conversation script, as illustrated in Table 2, con-
sisted of four lines for the conversant who initiated the
conversation and three lines for the child. Each conversa-
Table 1 tion started with a question for the child and ended with
an answer to the child’s last question. The middle of the
Conversation Topics per Participant conversation consisted of questions and answers between
the conversant and the child. For the three responses
Conversation Topic available for the child to make, each was considered cor-
Gary rect only if the child provided an appropriate answer and
A blue, red, and green yo-yos follow-up question.
A’ black and red yo-yos
Videos. VHS videotapes were used for video modeling.
B Big Bird and Bert
For each conversation topic (i.e., topics A through D),
C car and jet transformers the experimenter randomly picked and videotaped three
D robot and coloring book versions of conversations (e.g., A1, A2, A3; B1, B3, B5,
Connor etc.). [Note: For the remainder of this article, each set of
A robot and coloring book three conversations will be referred to as a separate video.
These videos consisted of conversations that were clearly
A’ monster and typewriter
modeled by adults. To facilitate generalization across
B Big Bird and Bert
conversants, many different models were used.]
C car and jet transformers
D blue, red, and green yo-yos

50 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Table 2

Examples of a Conversation Script Used for Connor (Robot and Coloring Book)
Conversation D1 Conversation D4
CONVERSANT: What color is the robot? CONVERSANT: Is the robot tall?
CHILD: Red, blue, yellow, and white. What color are his hands? CHILD: Yeah. Does he talk?
CONVERSANT: Blue. What is he holding? CONVERSANT: No. Is he smiling?
CHILD: A coloring book. Does he color? CHILD: Yes. Do his arms move?
CONVERSANT: No. Do you like to color? CONVERSANT: Yeah, watch. Do you want to shake his hands?
CHILD: Yes. Can we color a picture? CHILD: Sure. How does he move?
CONVERSANT: Yeah, great idea! CONVERSANT: He rolls. Look!

Conversation D’ Conversation D5
CONVERSANT: What color is the monster? CONVERSANT: Does Ralph have feet?
CHILD: Blue. What color are his hands? CHILD: No. How does he walk?
CONVERSANT: Silver. What is he holding? CONVERSANT: He rolls. Does he have hands?
CHILD: A typewriter. Does he type? CHILD: Sure does. Can he pick things up?
CONVERSANT: No. Do you like to type? CONVERSANT: Yes. Would you like to see?
CHILD: Yes. Can we type? CHILD: Yes. Can he pick up the book?
CONVERSANT: Yeah, great idea! CONVERSANT: That’s easy.

Conversation D2 Corresponding D-Test Conversation

CONVERSANT: What is that toy? CONVERSANT: Is that a robot?
CHILD: A robot. What’s his name? CHILD: Yes. Is the robot cute?
CONVERSANT: Ralph the robot. What color are his ears? CONVERSANT: Yes. Is the robot big?
CHILD: Red. Do they move?
CONVERSANT: Yes, they grow. Do you want to see?
CHILD: Sure. May I pull his ears?
CONVERSANT: Sure.

Conversation D3
CONVERSANT: Is the robot holding something?
CHILD: Yes. What is it?
CONVERSANT: A coloring book. Do you want to see some
pictures?
CHILD: Yes. Are there pictures of robots?
CONVERSANT: No. What’s in the picture?
CHILD: A dog. Can I fill in his ears?
CONVERSANT: Yeah. That would be fun.

Setting Baseline and probe conditions were conducted in the

treatment setting and two generalization settings. Gen-
Video modeling sessions were conducted two afternoons eralization settings consisted of a park-like play yard and
per week in one of two (2.9m x 2.9m) therapy rooms the participants’ home. Conversations with the conver-
at the after-school program the children attended. Each sant and probes with a peer with autism were conducted
room contained a table, chairs, a cabinet for toys, an in- in the treatment rooms. Probes with the child’s sibling
tercom for audio access, and a one-way mirror for view- were conducted at the child’s home. Finally, probes with
ing. These rooms are referred to as the treatment rooms. a typically developing peer were conducted at an out-
During the video modeling procedures, a VHS video doors play area lawn.
cassette player, the stimuli, and a television monitor were
also located inside the treatment rooms.

JSET 2008 Volume 23, Number 3 51

 Journal of Special Education Technology

Design ample, the child was praised for paying attention and
sitting nicely while working. However, at no time was
A multiple baseline design across subjects was used. Ad- the child praised directly for conversing.
ditionally, a multiple baseline design within subjects
across Conversations A and B was used to determine the Video modeling. Video modeling was presented when the
effects of video modeling on response variation in con- children had completed a predetermined number of con-
versational speech, and a multiple probe design across versations with Conversation A scripts.
person, settings, and stimuli was used for Conversations
C and D. The order of the presentation of conversations During video modeling sessions, the experimenter pre-
for each participant varied to reduce the likelihood of sented three videotaped versions of Conversation A (e.g.,
order effects. Thus, the conversation topic A for Gary A1, A2, A3 or A2, A4, A5). During the presentation of
was about yo-yos while Conversation A for Connor was the videotapes, the child sat next to the experimenter
about a robot and a coloring book (see Table 1). facing the television. The child was instructed to watch
the television and listen to what the models were talk-
Video modeling was presented when the children had ing about. Once the video was finished, the experimenter
completed a predetermined number of A-test and B-test said, “Now we’re going to talk about the toys.”
conversations. Gary had five A-test and six B-test conver-
sations, whereas Connor had seven A-test and eight B- The experimenter then conducted three conversations
test conversations. After video modeling was shown for using Conversation A-test. Prior to video modeling, the
Conversation A, the children were tested using Conver- amount of response variation across A-test Conversations
sation A-test. Once criterion had been met on Conversa- of each condition was assessed. Thus, variation during
tion A-test, probes for generalization were taken. baseline for Conversation A, across each A-test with the
experimenter, was assessed. If the child demonstrated a
Procedure greater amount of response variation across the three A-
tests than across the A-tests with the experimenter dur-
Baseline. The conversant and the child sat facing each ing baseline, the experimenter continued with the A-test
other, holding the appropriate stimuli. The conversant conversations.
began by asking a question about the stimuli (toys) from
the prescribed script. The child was expected to answer A-test conversations were conducted until the child had
the question and ask a contextually appropriate question completed as many A-test conversations with the experi-
within 10 seconds. If the child did not ask a question, menter after video modeling as the child had completed
the conversant recorded an incorrect response and con- during baseline. Thus, Gary had five A-tests with the
tinued with the script. If the child answered the origi- experimenter in baseline and five A-tests with the ex-
nal question and asked another question, the conversant perimenter after video modeling. Likewise, Connor had
responded by answering the child’s question and asking seven A-tests with the experimenter during baseline and
another question. This procedure continued until the seven A-tests after video modeling.
child had been presented with the opportunity to com-
plete three responses (three lines). The conversation was Criterion was met if the child made three times the num-
then closed with the response to the child’s last question ber of varied responses than he made during baseline. If
or with an appropriate ending response. the child did not meet criterion on Conversation A, vid-
eo modeling procedures were repeated. If the child met
During the conversation, the toy was given to the child criterion on Conversation A, he was tested across all of
if he asked for it. Thus, the toy served as a natural rein- the generalization probes. If the child failed to generalize,
forcer for the child’s question-asking. Also, the conver- video modeling was shown for Conversation B, and the
sant provided praise such as “uh-huh” or “right” after the procedure described above was followed. Once the child
child’s answer to encourage the child to continue with met criterion, he was again tested for generalization.
the conversation. After the conversation was completed,
reinforcement, such as tokens, snack, break, or praise, Generalization probes. Every probe session was conducted
was given for behaviors other than conversing. For ex- using the same procedures as those described in baseline.

52 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

During the probe sessions the conversant was the experi- tions were used to assess the generalization of conver-
menter, a typically developing peer, or a sibling. sational skills to topics that had not yet been modeled.
After video modeling and after the criterion for Con-
Generalization across persons. These probes were conduct- versation A had been met, the experimenter tested for
ed to test for generalization across persons with another generalization across the other topics of conversation
child with autism. The conversation was based on the (Conversations B-D).
robot and coloring book, which was Conversation D
for Gary and Conversation A for Connor (see Table 1). Prompted probes. For each conversation (A-D), the exper-
Based on the flip of a coin, Gary was determined to be imenter conducted special prompted probes during base-
the initiator of the conversation. To let Gary know that line only for test conversations. The experimenter began
he was to start the conversation, the experimenter said, the conversation just as in baseline, but if the child was
“Gary, you start and talk to Connor.” If he did not begin not asking a varied question, the experimenter instructed
the conversation, the experimenter instructed him to do the child to, “Say something different.” These probes
so by saying, “Say, is that a robot?” It was expected that were presented prior to video modeling to ascertain that
Gary might not know how to initiate the conversation the child could not learn response variation merely by
since his role, as indicated in the video, was to respond to being instructed to do so.
the conversation upon the initiation. Further, if a child
failed to provide a conversational line, he was given a
prompt so that the other child would be able to respond
Ancillary Data
and continue the conversation. All prompted lines were Variation in question-asking. The children were video-
scored as incorrect. taped bi-monthly, in a 10-minute free play situation.
The child was either with another child with autism, a
Generalization across persons and settings. Two different typically developing peer, or a therapist from the after-
probes were used to assess generalization across persons school program (these persons were not associated with
and settings together. The experimenter conducted probes the study). Transcripts of the children’s speech were used
with a typically developing peer and the child outside in to calculate the number of different questions the child
a park-like play yard. The typically developing peer pre- asked during these sessions. The number of different
sented three A-test conversations and three B-test con- questions the child asked during baseline with each of
versations, just as the experimenter did in baseline. The these persons was compared with the number of different
second type of probe, which assessed generalization across questions the child asked these persons following video
persons and settings, was conducted with the child’s sib- modeling. This was done to assess the effects of video
ling at home. Both children had siblings who presented modeling on the children’s use of question-asking in an
three A-test conversations and three B-test conversations, unstructured social situation.
just as the experimenter did in baseline. In both families
the child’s mother was taught to collect the data. Social behaviors. The same 10-minute free play situa-
tions were scored for the frequency of eight social be-
Generalization across new stimuli. Conversation A' was haviors (see Table 3). The frequency of these behaviors in
used to assess generalization to new stimuli. For each baseline was compared with their frequency after video
Conversation A', the stimuli and the lines used were modeling. Social behaviors were assessed to determine if
similar to those in the corresponding A Conversation. learning variation in conversations had an effect on these
For example, for Gary, Conversation A was about a red/ social skills.
blue yo-yo and a green yo-yo. Conversation A' was about
a black yo-yo and a red yo-yo. For Connor, Conversation
A was based on a robot and a coloring book, whereas
Dependent Measures
Conversation A' was about a monster and a typewriter. All responses during conversations were initially record-
ed as correct or incorrect. The criterion for a correct re-
Generalization across topics of conversation. A multiple- sponse was providing a contextually appropriate answer
baseline design across A- and B-test conversations, and and asking a contextually appropriate question per each
a multiple-probe design across C- and D-test conversa- line of the conversation. The line could be one modeled

JSET 2008 Volume 23, Number 3 53

 Journal of Special Education Technology

the target behavior. So, once the requirement of answer-

Table 3
ing and asking a question was met, and the response was
deemed correct, it was determined whether the response
Operational Definitions of Social Behaviors
met the definition of a varied response.
Social Behavior Definition
Variation was determined by the content of the question
With others Child is physically within two
feet of a person(s).
provided by the child in each line of the conversation.
Variation of speech was recorded according to the num-
Parallel play Child plays with a toy(s)
within two feet of another
ber of novel appropriate responses the child made in his
person who is playing in the three lines. Thus, the child had an opportunity to provide
absence of verbal and nonver- three varied responses per conversation. For a response to
bal interactions with others. be recorded as varied, the entire line must meet the defi-
Approach Child walks (moves) within nition of a correct response (consist of both an answer
two feet of a person in the and a question) and be of contextually appropriate but
absence of a cue from others different content than any other line previously said in
and stays with the person for
the condition (during baseline or after video modeling).
10 seconds or more.
For example, if the child said, “ May I play with this?”
Avoidance Child walks away (two feet
regarding the blue car in the first line of the conversation,
or more) when approached
by another person, or resists and then asked “May I play with the blue car?” in some
an interaction (e.g., trying to later line of a conversation after video modeling, a score
get out of a hug, or getting off of zero variation for the line would be given, because
someone’s lap). in both instances the child expressed the same content
Interactive play Child plays with a person by (playing with the blue car). However, if the child asked
exchanging toys, playing with to “drive” the blue car, he would be credited with two
the same toy(s) at the same varied responses within the conversation because playing
time.
and driving express varied content. If, on his third line,
Nonverbal response Child provides an appropri- the child, still talking about the blue car, asked to “hold”
to others ate nonverbal behavior in
response to a person’s verbal
the car, this too would be considered varied content, and
or nonverbal behavior while the child’s conversation would get a score of 3/3 varied
standing within two feet of the responses. This is an example of a change in content in
individual or while maintain- terms of verb but not object (same blue car). Another
ing eye contact. example would be if the child asked to play (same verb)
Nonverbal social Child begins an activity or with varied objects. Finally, and hopefully, variation in
initiation affection behavior, within both verb and object would occur (e.g. “The sharks eat
two feet of another person or
pizza”; I like cheese pizza”; “Sharks really eat fish”).
while maintaining eye contact,
in the absence of a cue from
others. The three lines in the conversations, the within-con-
Social interaction The other person starts an versation variation measure, was a content-comparative
directed at child interaction while standing measure of the three lines. Since variation is a relative
within two feet of the child or concept, there could be two lines that varied from each
orients his/her body toward other or three lines that varied from each other. One line
the child, or maintains eye could not vary from another without the other line being
contact with the child.
counted as varied as well. Thus, there was no way a con-
versation could have one varied response. The minimum
varied responses would be two, and the maximum, three.
Therefore, for each conversation, the child could only re-
in the video or a novel one. Thus, each child had an op- ceive a score of zero, two, or three varied conversational
portunity to make three lines or three correct responses lines. This provides stringent scoring criteria for response
per conversation. However, variation in conversation was variation to avoid inflation of results.

54 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Interobserver Reliability and 4 show the same data for Connor. Figures 5 and
6 (Gary) and 7 and 8 (Connor) show the cumulative
Interobserver agreement was assessed for variation in number of varied responses not within conversation, but
conversation data, transcripts of free play sessions for within condition (baseline, video modeling, post-video
question-asking data, and frequency of social behaviors modeling) for each conversation across the study.
during free play data. This was done for both children in
all conditions. Interobserver agreement was determined Overall, the results show an increase in variation in
by calculating the number of agreements divided by the conversational responses for both boys within conversa-
total number of agreements plus disagreements multi- tions after video modeling. Additionally, generalization
plied by 100. occurred across the probe conditions and to untrained
conversations. Cumulative number of varied responses
Agreement among observers was calculated as 37% for also increased for both boys. Ancillary data suggest an
the conversational data for Gary and 40% for Connor. increase in question-asking during free play time as well
During the conversations, the second observer either sat as increases in a few social behaviors. The results for each
behind a one-way mirror, in an unobtrusive location in child are discussed below.
the experimental room, outside, or at the child’s home.
For probes taken at the child’s home, the mother was
trained as the reliability observer. The experimenter com- Within-Conversation Variation
pared the number of correct responses and the number Gary. Gary seldom showed conversational variation
of varied responses recorded by each of the observers for across the conditions for Conversation A during baseline
the conversations. Interobserver agreement for both chil- (see Figure 1). Two probes were observed when he var-
dren was 100%. ied his speech for the A-test Conversation; otherwise, the
majority of his data showed no variation during Conver-
Transcript data were calculated for 41% of all free play sation A and A-test probes during baseline. Prompting
sessions across all conditions for Gary and 37% for (“Say something different”) did not facilitate conversa-
Connor. Transcripts were assessed using word-for-word tion variation. Gary needed only two presentations of
agreement. A disagreement was recorded if the primary video modeling before he met criterion of tripling his
rater had recorded a word that the secondary rater did variation within conversation compared with baseline.
not record and vice versa. Interrater agreement for the Generalization probes with a typically developing peer
transcripts was 90% for both children. and a peer with autism suggested partial generalization.
Social behaviors during free play sessions across all con- Generalization across stimuli, which was assessed through
ditions were determined for 40% of all sessions for Gary Conversation A', did not show increased variation after
and 38% of all sessions for Connor. Observers were video modeling was presented for Conversation A. While
trained by reviewing the operational definitions for each it is unclear whether there is a strong collateral effect of
social behavior, reviewing and scoring a tape with the video modeling for Conversation A upon Conversation
experimenter, and then scoring a tape independently to C, there appears to be an effect for Conversations C and
assess reliability. Interobserver agreement was assessed for D (see Figure 2).
each behavior during each session, and total agreement
was calculated. Interobserver agreement was 96% for Connor. Figure 3 shows the effect of video modeling on
Gary and 98% for Connor. within-conversation variation for Connor. During base-
line, there were few conversations during Conversation A
(top panel) in which variation in conversation occurred.
However, after two presentations of video modeling,
Results variation in within-conversation increased and met cri-
The results of video modeling on conversation varia- terion of triple the frequency of baseline. Generalization
tion for Gary and Connor are illustrated in Figures 1-6. probe data did not yield positive results for Conversa-
Specifically, Figures 1 and 2 show the number of varied tion A. However, there did appear to be improvement
responses within conversations for Gary while Figures 3 in conversation variation for both Conversation A' and

JSET 2008 Volume 23, Number 3 55

 Journal of Special Education Technology

Figure 1

Number of varied responses for Gary within conversations for Conversations A, A', and B with
experimenter, peer with autism, and typical peer
Gary
Conversation A Video Modeling
3 Baseline Conversation with
Exp.

Peer with Test with

1 Exp. Typical Peer Denotes Prompted Conversation
Autism

***
0 ***

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120
Number of Varied Reponses

Conversation A`
3

2
Conversation with
Exp.
1

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversation B
3

2 Conversation with
Exp.
Typical
Peer Denotes Prompted Conversation
1 Test with Exp. Peer with Autism

0 ***

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversations

Conversation B. Data collected for Conversations C and Gary. Gary showed little variation across conversations in
D show erratic performance (see Figure 4). each condition during baseline, and after five A-test con-
versations with the experimenter, he had only asked three
varied questions. By comparison, after one presentation
Cumulative Variation in Responses Across of video modeling, he made nine varied responses. When
Conversations video modeling was presented a second time, Gary asked
Since there are limits to how increases within conver- an additional seven new questions. In addition, he gen-
sation in response variation can be measured, it is also eralized variation across people and settings, and the data
important to determine the cumulative amount of varied showed an increase in variation across topics of conversa-
conversational responses the children learned after video tion (see Figures 5 and 6).
modeling. As seen in Figures 5-8, both Gary and Con-
nor made large increases in the cumulative frequencies Connor. Connor, showed little variation across conversa-
of varied and novel responses after video modeling com- tions in each conditions for Conversations A and B (see
pared with the number of different responses they made Figures 7 and 8). After video modeling was presented,
during baseline in each condition. however, he made 8 novel and varied responses and ob-

56 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Figure 2

Number of varied responses for Gary within conversations for Conversations C and D with experimenter,
peer with autism, and typical peer

Gary
Conversation C
3

Conversation Sibling
1
Typical with Exp.
Number of Varied Reponses

Peer
0

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversation D
*
3

* ***
2

Conversation
with Exp. Test with Exp.
1

** ** Denotes Prompted
0
Conversation

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversations

tained a cumulative number of 16 varied responses after Table 4

the second presentation of video modeling. Generaliza-
tion occurred across Conversations B-D. However, more Response Variation During Free Play Sessions
varied responses did not occur in the probes for general-
ization across persons or settings. Post-Video
Condition Baseline Modeling
Gary
Question-Asking During Free Play
Therapist 0/1 3/4
The number of varied questions out of the total number Peer with Autism 0/0 0/0
of questions asked during the 10-minute free play ses- Total 0/1 3/4
sions are listed in Table 4.
Connor

Gary. During baseline, Gary asked one question during Therapist 3/3 16/18
his free play sessions with a therapist and with a peer with Peer with Autism 1/2 8/18
autism, and this question occurred with the therapist. Typical Peer 2/2 4/4
However, since there was no other question to compare Total 6/7 28/40
it to, this question could not be counted as varied. Af-
ter video modeling, Gary asked four questions, all in the Note. The scores represent ratios between the number of varied
therapist condition; three of them were varied responses. responses and the total number of questions asked.
Thus, Gary increased his variation in question-asking

JSET 2008 Volume 23, Number 3 57

 Journal of Special Education Technology

Figure 3

Number of varied responses for Connor across conversations for Conversations A, A', and B with
experimenter, peer with autism, typical peer, and sibling
Connor
Conversation A
Baseline Video Modeling

3 Typical Peer

2
Conversation with
Exp.
1 Test with Exp. Peer with Autism

*** Denotes Prompted

0
Conversation

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120
Number of Varied Reponses

Conversation A`
3

1 Conversation with
Exp.

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversation B
3

Conversation with
1 Test with Exp. Typical Peer
Exp. Peer with Autism

0 *** Denotes Prompted

Conversation

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversations

after video modeling and the frequency of his question- Social Behaviors During Free Play
asking increased, but he did not ask a question to his
peer, and his overall number of questions was not high. Figures 9 and 10 show the eight social behaviors mea-
sured for both children in all conditions for baseline and
Connor. Connor displayed more question-asking behav- after video modeling. Changes in the social behaviors did
ior during baseline than Gary, speaking across all condi- not reveal any consistent findings for the two boys.
tions and varying his responses for six out of his seven
questions. Importantly, after video modeling, Connor
dramatically increased the frequency of his questions
to 40 questions, 28 of which were considered varied Discussion
responses. In the present study, two boys with autism were taught
variation in conversational speech using video model-
ing technology. Scripts of various topics of conversations
were written about a number of toys that were the ba-

58 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Figure 4

Number of varied responses for Connor within conversations for Conversations C and D with
experimenter, peer with autism, and typical peer

Connor
Conversation C
3

2
Denotes Prompted
Conversation
Number of Varied Reponses

Conversation
1 with Exp. Test with Exp.

0 *** ***

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversation D
3

*
2
Denotes Prompted
Test with Exp.
1 Conversation Peer with Autism Conversation
with Exp.
*** **
0

-1
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 106 113 120

Conversations

sis of discussion. For each toy topic, several scripts were study was to examine whether variation in the children’s
written in order to provide examples of response varia- responding could be increased beyond the modeled
tion. This form of multiple-exemplar training has been phrases depicted in the videos. The results of the present
shown to increase response generalization (e.g., Charlop- study also show the efficacy of the multiple conversation
Christy & Daneshvar, 2003; Stokes & Baer, 1977). presentation format as a means to enhance variation in
the conversation of the children with autism. Further,
The varied conversations were videotaped and presented generalization probes suggested both stimulus and re-
to the children via a video modeling protocol devised by sponse generalization.
Charlop and Milstein (1989) and replicated many times
in the literature (cf. Charlop-Christy, Le, & Freeman, The present research adds to the literature in several
2002). Generalization was assessed across untrained but ways. First, the findings provide additional evidence for
similar toys, settings, typically developing peers, and the effectiveness of video modeling in general (Charlop-
peers with autism. Additionally, response generalization Christy, et al., 2000) as well as for conversational speech
to other topics of conversations based on different toys specifically (Charlop & Milstein, 1989; Scherer et al.,
was assessed. 2000).

The results replicated the previous video modeling study Second, the study shows how a multiple exemplar (mul-
by Charlop and Milstein (1989) showing rapid effects tiple conversation) training paradigm (Stokes & Baer,
of video modeling on teaching conversational speech af- 1977), in this case the presentation of multiple conversa-
ter long baselines. However, the purpose of the present tions about the same topics in the video, can facilitate

JSET 2008 Volume 23, Number 3 59

 Journal of Special Education Technology

Figure 5

Cumulative variation across test conversations for Gary for Conversations A, A', and B
Gary
Conversation A Video Modeling
18 Baseline
16
14
12
10
8 Test with Peer with
6 Exp. Autism
Typical Peer
4
2
0
-2
Cummulative Varied Reponses

1 10 19 28 37 46 55 64 73 82 91 100 109 118

18 Conversation A`
16
14
12
10
8
6 Test with
Exp.
4
2
0
-2
1 10 19 28 37 46 55 64 73 82 91 100 109 118

18 Conversation B
16
14
12
10
8
Test with Peer with
6 Exp. Typical Peer Autism
4
2
0
-2
1 10 19 28 37 46 55 64 73 82 91 100 109 118

Conversations

response variation, response generalization, and stimulus with autism, a typically developing peer, and a sibling at
generalization. These results are encouraging in that they home.
demonstrate the effectiveness of a teaching technology,
video modeling that has been deemed cost efficient as Fourth, the inclusion of ancillary measures of social
well as expedient, in teaching a difficult to teach behavior behaviors allowed for assessment of any concomitant
to children with autism. Charlop-Christy and Daneshvar changes that may be seen in association with video mod-
(2003) as well as LeBlanc and colleagues (2003) demon- eling. The observations of social behaviors during free
strated the utility of incorporating multiple exemplars for play conditions before and after video modeling allowed
teaching a perspective-taking task. This may be necessary for such an assessment. Also, the transcription of record-
when teaching more abstract concepts such as conversa- ed free play sessions before and after video modeling al-
tional variation. lowed for a comparison of the number of questions asked
before versus after video modeling.
Third, the present research tested the children under
naturalistic conditions in which variation in conversa- Finally, in addition to the extensive analysis of both stim-
tional speech would be most desirable; with another peer ulus and response generalizations, the dependent variable
was measured in two ways. First, we assessed variation

60 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Figure 6

Cumulative variation across test conversations for Gary for Conversations C and D
Gary
18
16 Conversation C
14
Test with
12
Exp.
10
8
6
4
2
0
Cumulative Varied Reponses

-2
1 10 19 28 37 46 55 64 73 82 91 100 109 118

18
Conversation D
16
14
12
10
8 Test with
Exp.
6
4
2
0
-2
1 10 19 28 37 46 55 64 73 82 91 100 109 118

Conversations

in conversational speech within conversation. While or whether they understood how to take one line in one
the data showed positive results, this approach limits conversation (say, a line from Conversation C), modify
what we can assess regarding what the children learned. it, and apply it in the context of the toys and topic in
In the present study, as in previous research (Scherer et Conversation A-test.
al., 2001), the dependent variable had a predetermined
number of responses to score. That is, for each conver- This could be measured by the cumulative data figures.
sation there was a maximum of three correct responses. From this data analysis, we learned that the children not
Additionally, due to the nature of content variation, only learned through video modeling and met criterion
we did not allow for one correct line to be counted as in their within conversation variation but they also great-
a correct varied response unless there was at least one ly increased variation in responding after video model-
other response to which its content could be compared. ing across the study. From this cumulative analysis, we
So, our range of within conversation correct varied re- were also able to anecdotally observe some qualitative
sponses was zero, two, and three responses. While this changes in the boys’ varied responses. For example, while
measure can determine the efficacy of video modeling, questions tended to be short and basic during baseline,
and improvement over baseline, it may limit what we can toward the end of the study, the children’s questions be-
know about what the children learned about response came more sophisticated. For example, Connor asked,
variation as a concept. That is, it prevents us from deter- ”Where is Robinson Crusoe from?”
mining whether the children acquired the “learning to
learn” concept (Schreibman, Charlop, & Koegel, 1982)

JSET 2008 Volume 23, Number 3 61

 Journal of Special Education Technology

Figure 7

Cumulative variation across test conversations for Connor for Conversations A, A', and B
Connor
Conversation A Video Modeling
16 Baseline
14
12
10
8 Typical Peer
6 Test with Exp. Peer with Autism
4
2 * Denotes Prompted
** Conversation
0
-2
Cumulative Varied Reponses

1 15 29 43 57 71 85 99 113 127 141

Conversation A`
16
14
12
10
8
6
4
2
Test with Exp.
0
-2
1 15 29 43 57 71 85 99 113 127 141

16
Conversation B
14
12 Denotes Prompted Conversation
10
Typical Peer
8
6 Peer with Autism
Test with Exp.
4
***
2
0
-2
1 15 29 43 57 71 85 99 113 127 141

Conversations

The results of the cumulative analysis also provide ad- entire conversation, including the experimenter’s roles.
ditional ammunition against arguments that mimicking This, indeed, suggests that variation in conversation as
video presentations of behaviors may merely be imitation well as varied initiations were learned. Unfortunately, we
and rote learning. While this line of criticism is less ap- had only one dyad of children with autism. We felt that
plicable here, where we found such variation in response, to switch off and change the boys around to take turns
even within conversation, the cumulative analysis does being initiators might have been too much variation for
provide evidence of continual engagement of the desired them. However, it is possible that Connor could have
behavior in incarnations never modeled. Another piece fared just as well as Gary.
of evidence is offered as well. Recall that when being
tested for conversation-variation acquisition, the con- Our results support the fast and efficacious findings of
versations for each child with autism were initiated by video modeling and showed the applicability to con-
the experimenter. However, when speaking with a peer versation variation. However, some of the results show
with autism, one child had to assume the alternate role; that we need to continue to explore modifications in
that of the experimenter. Gary demonstrated that he was video modeling to enhance its effects. The generaliza-
capable of engaging in conversational variation for the tion effects in the present study were not as positive as

62 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Figure 8

Cumulative variation across test conversations for Connor for Conversations C and D
Connor
16 Conversation C
14
12 Denotes Prompted
10 Conversation
8
6
Cumulative Varied Reponses

4 **
** *
*
2
Test with Exp.
0
-2
1 15 29 43 57 71 85 99 113 127 141

16
14 Conversation D
12 Denotes Prompted
10 Conversation
8
6 **
Peer with Autism *
4 *
*
2 *
Test with Exp.
0
-2
1 15 29 43 57 71 85 99 113 127 141

Conversations

in other video modeling studies. This may be due to (a) suggests that the varied conversation format might have
the nature of the target behavior, that being an abstract enhanced the variation of the children’s speech in other
concept; (b) the nature of autism, that being difficult to situations, specifically the number and variation in the
obtain generalizable effects; or (c) limitations within our questions they asked.
procedures. The inclusion of only two participants sug-
gests the need for replication across more dyads of chil- This is an important ancillary gain for several reasons.
dren. Further, the ancillary data showed mixed results. First, children with autism tend to be responders and
The free play observations of eight social behaviors did not question askers (Schreibman, 2005). Therefore, an
not show anything consistent with the children in the important behavior was increased, likely due to the in-
various conditions during baseline and after video mod- tervention. Once children with autism start asking more
eling. Perhaps variation in conversations has little to no questions, they will likely have more interaction with the
effect upon other social behaviors or perhaps such effects environment and, therefore, obtain more information.
need more time to show up. There are a number of pos- Also, as mentioned earlier, children with autism tend to
sibilities but there were no immediate effects. However, say the same phrases or ask the same questions repeatedly
immediate and dramatic collateral effects were seen with and perhaps perseveratively (e.g., Kanner, 1943; Rim-
regard to question-asking during free play. Perhaps this land, 1964). This is considered one of the inappropriate
is due to the more relevant nature of the behaviors. The behaviors of autism (APA, 2000). Thus, with the increase
number of questions and, more importantly, the variety of question variation, the children are not only access-
of questions asked increased after video modeling. This ing the environment and the information in it, but also

JSET 2008 Volume 23, Number 3 63

 Journal of Special Education Technology

Figure 9

Frequency of social behaviors for Gary during free play sessions

Gary Gary

100 Baseline 100

Baseline
Video Modeling
Video Modeling
80 80

Percent Occurrence
Percent Occurrence

60 60

40
40

20
20

0
With Others Parallel Play Approach Avoidance Interactive Nonverbal Nonverbal Social 0
Play response to social interaction With Others Parallel Play Approach Avoidance Interactive Nonverbal Nonverbal Social
others intiation directed at Play response to social interaction
child others intiation directed at
With Therapist
child
100 With Typically Developing Peer
100

80 80

60 60
Percent Occurrence
Percent Occurrence

40 40

20
20

0
0 With Others Parallel Play Approach Avoidance Interactive Nonverbal Nonverbal Social
With Parallel Approach Avoidance Interactive Nonverbal Nonverbal Social Play response to social interaction
Others Play Play response to social interaction others intiation directed at
With Peer with Autism others intiation directed at Total child
child

decreasing a perseverative behavior (Charlop & Milstein, References

1989).
Apple, A. L., Billingsley, F., & Schwartz, I. S. (2005). Effects of video
modeling alone and with self-management on compliment-giving
In conclusion, the present study extends the literature behaviors of children with high-functioning ASD. Journal of Posi-
by providing additional adjustments, analyses, and de- tive Behavior Interventions, 7, 33-46.
pendent variables when incorporating video modeling American Psychiatric Association. (2000). Diagnostic and statistical
techniques into intervention strategies for children with manual of mental disorders (4th ed., text rev.). Washington, DC:
autism. The results suggest that video modeling can be Author.
an effective technology for teaching children with autism Bellini, S., & Akullian, J. (2007). A meta-analysis of video model-
to vary their conversational speech in order to speak with ing and video self-modeling interventions for children and ado-
lescents with Autism Spectrum Disorders. Exceptional Children,
several persons on a variety of topics. 73, 264-287.
Charlop-Christy, M. H., & Daneshvar, S. (2003). Using video mod-
eling to teach perspective taking to children with autism. Journal
of Positive Behavior Interventions, 5, 12-21.
Charlop-Christy, M. H., & Kelso, S. E. (2003). Teaching children
with autism conversational speech using a cue card/written script
program. Education and Treatment of Children, 26, 108-127.

64 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Figure 10

Frequency of social behaviors for Connor during free play sessions

Connor Connor

100 Baseline
100 Baseline
Video Modeling
Video Modeling
80 80
Percent Occurrence

Percent Occurrence
60
60

40
40

20

20

0
With Parallel Approach Avoidance Interactive Nonverbal Nonverbal Social
Others Play Play response social interaction 0
to others intiation directed at With Parallel Approach Avoidance Interactive Nonverbal Nonverbal Social
child Others Play Play response social interaction
With Therapist to others intiation directed at
100 child
Total

80
Percent Occurrence

60

40

20

0
With Parallel Approach Avoidance Interactive Nonverbal Nonverbal Social
Others Play Play response social interaction
to others intiation directed at
child
With Peer with Autism

Charlop-Christy, M. H., Le, L., & Freeman, K. A. (2000). A com- Leblanc, L., Coates, M., Daneshvar, S., Charlop-Christy, M., Morris,
parison of video modeling with in vivo modeling for teaching C., & Lancaster, B. (2003). Using video modeling and reinforce-
children with autism. Journal of Autism and Other Developmental ment to teach perspective-taking skills to children with autism. Jour-
Disorders, 30, 537-552. nal of Applied Behavior Analysis, 36, 253-257.
Charlop, M. H., & Milstein, J. P. (1989). Teaching autistic children MacDuff, G. S., Krantz, P. J., & McClannahan, L. E. (1993). Teach-
conversational speech using video modeling. Journal of Applied ing children with autism to use photographic activity schedules:
Behavior Analysis, 22, 275-285. Maintenance and generalization of complex response chains. Jour-
nal of Applied Behavior Analysis, 26, 89-97.
Goldstein, H., & Cisar, C. L. (1992). Promoting interaction during
sociodramatic play: Teaching scripts to typical preschoolers and Matson, J. L., & Sweezy, N. (1994). Social skills training with autistic
classmates with disabilities. Journal of Applied Behavior Analysis, children. In J. Matson (Ed.), Autism in children and adults: Etiol-
25, 265-280. ogy, assessment, and intervention (pp. 241-260). Pacific Grove, CA:
Brooks/Cole.
Kanner, L. (1943). Autistic disturbances of affective contact. The Ner-
vous Child, 2, 217–50. Nikopoulos, C. K., & Keenan, M. (2003). Promoting social initia-
tion in children with autism using video modeling. Behavioral In-
Krantz, P. J., & McClannahan, L. E. (1993). Teaching children with
terventions, 18, 87-108.
autism to initiate to peers: Effects of a script-fading procedure.
Journal of Applied Behavior Analysis, 26, 121-132.

JSET 2008 Volume 23, Number 3 65

 Journal of Special Education Technology

Nikopoulos, C. K., & Keenan, M. (2004). Effects of video modeling Stokes, T. F. & Baer, D. M. (1977). An implicit technology of gener-
on social initiations by children with autism. Journal of Applied alization. Journal of Applied Behavior Analysis, 10, 349-367.
Behavior Analysis, 37, 93-96.
Rimland B. (1964). Infantile autism: The syndrome and its implica- Author Note
tions for a neural theory of behavior. New York: Appleton-Century-
Crofts. Marjorie H. Charlop is in the Department of Psychology, Clare-
Schreibman, L. (2005). The science and fiction of autism. Cambridge, mont McKenna College. Laura Gilmore is no longer at Claremont
MA: Harvard University Press. McKenna College. Gina T. Chang is in the School of Behavioral
and Organizational Sciences, Claremont Graduate University.
Schreibman, L., Charlop, M. H., & Koegel, R. L. (1982). Teaching au-
tistic children to use extra stimulus prompts. Journal of Experimental We would like to thank Janice Milstein and Marion Moore for
Child Psychology, 33, 475-491. their assistance.
Sherer, M., Pierce, K. L., Paredes, S., Kisacky, K. L., Ingersoll, B., Correspondence concerning this article should be addressed to
& Schreibman, L. (2001). Enhancing conversation skills in chil- Marjorie H. Charlop, Department of Psychology, Claremont
dren with autism via video technology: Which is better, “self ” or McKenna College, Claremont, California 91711. E-mail:
“other” as a model? Behavior Modification, 25, 140-159.
[email protected]

66 JSET 2008 Volume 23, Number 3

 Journal of Special Education Technology

Author Guidelines

Journal of Special Education Editorial Policy

JSET is a refereed professional journal that presents up-to-date information and opinions about issues, research, policy,
and practice related to the use of technology in the field of special education. JSET supports the publication of research
and development activities, provides technological information and resources, and presents information and discussion
concerning important issues in the field of special education technology to scholars, teacher educators, and practitioners.

Guidelines for Preparation and Submission of Manuscripts to JSET

The Journal of Special Education Technology (JSET) is an open submission journal that publishes three non-topical issues
and one topical issue a year. Topics appropriate for the journal include, but are not limited to:
• Discussion of issues and trends in the use of technology in the field of special education.
• Reports of experimental or applied research that deals with the use of technology or assistive devices for people
with disabilities.
• Analysis of policy developments at the state and federal levels that impact the use of technology for people with
disabilities.
• Description of program models that incorporate the use of technology in the education of people with disabilities,
the transition of people with disabilities, the leisure time of people with disabilities, etc.
• Description of teacher education programs that incorporate the use of technology in the education of future spe-
cial educators and/or others whose focus is the provision of services to people with disabilities.
• Reviews of the literature that focus on technology in special education.
All manuscripts are judged according to a blind review process by three members of the editorial board, with the final deci-
sion for publication made by the editor. The reviewers will be selected on the basis of their expertise in the area addressed
or the methodology used. The review process normally will take 12 to 16 weeks. The principal author only will be notified
concerning publication of the article.

To Submit a Manuscript to JSET, Please Adhere to the Following Guidelines

1. Five (5) printed copies of the manuscript following all guidelines of the American Psychological Association Publication
Manual (APA) (5th edition, 2001).
2. Manuscripts generally should be no longer than 35 pages in length for an article submission; 10-12 pages in length
for submission to one of the associate editor columns.
3. Preferred positions of tables and figures should be indicated in the manuscript.
4. Authors are encouraged to write in “person-first” language: the person precedes the disability–for example, “people
with disabilities” not “the disabled.”
5. A cover letter should accompany the manuscript indicating that the manuscript has not been published in whole or a
substantial part by another publisher and that the manuscript is not currently under review by another journal.
6. After a manuscript has been accepted for publication and after all changes have been made, the authors must send a
copy of the manuscript and a computer file to the editor.

Send Manuscripts to or for Further Information Contact

J. Emmett Gardner, JSET Editor
Department of Educational Psychology, Special Education Program, University of Oklahoma
820 Van Vleet Oval, Room 326
Norman, OK 73019
voice: 405-325-1533 fax: 405-325-7661
email: [email protected] or [email protected]
View publication stats