Increasing communication skills: a case study of a man with autism spectrum disorder and vision loss.
Article Type: Brief article
Subject: Mentally ill (Education)
Mentally ill (Research)
Sign language (Study and teaching)
Sign language (Research)
Low vision
Authors: Kee, S. Brian
Casey, Laura Baylot
Cea, Clayton R.
Bicard, David F.
Bicard, Sara E.
Pub Date: 02/01/2012
Publication: Name: Journal of Visual Impairment & Blindness Publisher: American Foundation for the Blind Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2012 American Foundation for the Blind ISSN: 0145-482X
Issue: Date: Feb, 2012 Source Volume: 106 Source Issue: 2
Topic: Event Code: 310 Science & research
Accession Number: 282068142
Full Text: According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; American Psychiatric Association, APA, 2000), autism is a neurodevelopmental disorder that is characterized by impairments in social and communicative behaviors with great variations in ability, depending on developmental level, intelligence, and chronological age. Thus, autism is a spectrum disorder that affects a large number of children (1 in every 150) to various degrees (Centers for Disease Control and Prevention, 2007). Autism is further defined as qualitative impairments in social interaction, which include impairments in nonverbal and verbal behaviors ranging from the lack of spoken language with no attempt to use alternate modes to only mild delays in conversation (APA, 2000). Although the degree of communication problems varies, one common thread for all individuals who are diagnosed with autism is that if interventions are not designed to meet the individual's needs in a socially significant manner, the individual has a poor prognosis (Heward, 2009). Thus, it is important to consider the whole person when designing interventions by taking into account additional disabilities as well as individual strengths.

Two of the more common tools for teaching communication skills to individuals with disabilities related to communication are the Picture Exchange Communication System (PECS; Bondy & Frost, 1994), which relies on the exchange of pictures representing a desired object, place, or event, and mand training (Rosales & Rehfeldt, 2007; Skinner, 1957). Skinner (1957), a behaviorist, coined the term mands, along with other verbal operants (tacts, intraverbals, and echoics), under the umbrella of verbal behavior that is based on the principles of applied behavior analysis. The term mand is defined as "a verbal operant for which the form of the response is under the functional control of motivating operations and specific reinforcement" (Cooper, Heron, & Heward, 2007, p. 30); that is, people request specific items or things only when they are motivated to do so. This motivation is often derived from a state of deprivation. Both methods have endured much success.

Sundberg and Partington (1998) provided basic recommendations for mand training: (1) establish that a motivating factor or an establishing operation (such as a state of deprivation) is in place prior to beginning the session; (2) consider those items that are motivating to the individual; (3) use only items to which the therapist can easily control access; (4) select only those items or activities that the person understands through some means of communication, such as imitation or pulling you; (5) use reinforcers that are easy to deliver; and (6) use reinforcers that can be delivered on multiple occasions (such as a sip of juice instead of an entire cup). The study presented here used these recommendations.


The primary goal of the study was to teach a nonverbal man who was blind and was diagnosed with autism to use sign language to mand for two highly desirable items: food and drink. However, because of the severity of the autism and vision loss, the use of mands was selected over PECS, which relies heavily on the ability to see the item that is depicted.

To begin the mand training, some preliminary work was required to get the participant to use correct American Sign Language (ASL). The correct use of ASL was required because of the individual's use of idiosyncratic signs that differed from ASL and were all similar to one another in form, which resulted in the caretakers' inability to discern what the individual was requesting. This lack of ability to discern the participant's wants and needs resulted in the desired item not being delivered and the man engaging in self-mutilation or other destructive behaviors until the desired object was provided.


Participant and setting

The participant in the study was a nonverbal 28-year-old man who was blind because of a rare retinal disease he contracted at age 12, and who had profound autism. This man had a long history of being severely aggressive toward his parents, staff, general caregivers, and himself. He was referred to a behavioral service provider after numerous failed placements at supportive living agencies, where he often was supported by therapists for up to 70 hours a week. Before the sessions began, approval was obtained to conduct the study from the Institutional Review Board of the University of Memphis.

The sessions were conducted in the participant's home and took place at an empty table with long bench-style chairs on opposite sides of the table. The participant sat across the table from the therapist (the first author) during all the sessions.


During both the baseline and intervention, the researcher (the first author) ensured that an establishing operation (in this case, mild deprivation) was present. The state of deprivation from food and drink was ensured by conducting sessions during the participant's lunchtime, which was consistently three hours after his morning meal. Once the establishing condition was established, the participant was allowed 30 seconds to emit a sign for his preferred item. Contingent upon the sign that was given by the participant, he was provided with a desired reinforcement item in the form of a quick bite of food or a sip of soda. If the participant did not emit any sign, no differential consequence was provided, and the next trial started. If at any time during the study the participant emitted a known escape response, such as hand waving, attempting to leave the table, grunts, or sounds of agitation, he was allowed to escape, and the sessions were discontinued briefly. Allowing for sessions to be discontinued was necessary because of the participant's history of aggression with behavioral interventions and the possibility of him harming himself.

Research design

The research design was an AB design with a maintenance phase to ensure that the participant did not lose the acquired skills once the intervention was withdrawn.

Baseline. During the baseline phase, a verbal prompt was used to signify the availability of both food and drink, since the participant was blind. The therapist started off the session with the following statement: "There is food and drink on the table." Both mands were trained in one discrete trial. During the baseline condition, if the participant gave the sign for "eat" or "drink," the behavior was reinforced with the corresponding reinforcer. If he gave no sign or an escape response, the statement, "There is food and drink on the table" was presented again. Thus, differential reinforcement of communicative behavior was used during the baseline, intervention, and maintenance phases. Following the analysis of the baseline data, it was evident that the participant was using the sign for drink but not truly desiring a drink, as was demonstrated by his manding for a drink, receiving the cup, then throwing the cup. Thus, the intervention described next was designed to help the participant better understand that the sign for drink would result in a sip of a preferred liquid and that the sign for eat would result in a portion of a desired edible.


Intervention. A touch-prompt intervention was selected on the basis of the overallocation of the signs for drink. The touch prompt consisted of the researcher lightly touching the participant's hand and lightly pushing it in an upward direction on a fixed ratio schedule of every five discrete trials. The hand-overhand technique was deemed intrusive for the participant, since the sign was in his repertoire. Selecting a touch prompt and using it on a fixed ratio 5 schedule also made it easy to withdraw the prompt. Fixed ratio schedules are ideal when the goal is not to increase dependence on a prompt and eventually to withdraw the prompt completely (Cooper et al., 2007). The goal of the fixed ratio 5 schedule, combined with the differential reinforcement of communicative behavior, was to distribute his mands more evenly on the basis of his needs and to ensure that he was truly manding for what he wanted and not simply using a sign that, in the past, had intermittently resulted in the desired item from the caregiver (see Figure 1).

Maintenance. The maintenance phase followed the intervention to determine whether the mand training was successful and whether the participant was now able to mand independently for desired objects using the correct signs, as is illustrated in Figure 2.

Interobserver agreement

Interobserver agreement was taken during each mand training session. It consisted of a trained observer taking data independently from the researcher. Each time the participant manded in the given interval, it was scored accordingly. At the end of every session, a comparative analysis was done to calculate the total agreement. Data on interobserver agreement were taken 40.63% of the time during mand training sessions. The average score was 96.45%, with a range of 83.33% to 100%.

Social validity

Social validity, as defined by Cooper et al. (2007), is the extent to which intervention procedures are deemed acceptable, significant changes are evident in the target behavior, and important and significant collateral behaviors are produced as a result of the intervention. To assess social validity, the researcher asked the parents and staff to respond to two primary questions before and after the study using a Likert scale of 0-5 with the following assigned scores: strongly disagree (0), disagree (1), neutral (2), agree (3) or strongly agree (5). The questions were "How well do you feel the participant communicates that he is in need of something?" and "Do you feel that you understand precisely what he is in need of at that moment?" The results for social validity are presented in Figure 3 (for the parents) and Figure 4 (for the staff).




During the baseline phase, the percentage of mands for "eat" occurred at a low level with a descending trend. The mean for "eat" was 7.21%, with a range of 0% to 12.5%. The percentage of mands for "drink" occurred at a midlevel with an ascending trend. The mean for "drink" was 53%, with a range of 31.03% to 68.97%. This ascending trend occurred even though the individual was receiving a cup with a preferred drink and throwing it as opposed to consuming it. The data indicated that the participant was not making a connection between manding for the preferred item of his choice and his actual receipt of that item.


Following the baseline phase, a touch prompt was given to guide the participant to the sign for "eat." The mean for manding "eat" rose to 31.15%, with a range of 25.71% to 40.63%, along with the actual consumption of the food that he was given on making the sign. Manding for drink remained at the same level as at the baseline, with a slightly ascending trend with a mean of 49.78% and a range of 37.5% to 61.54%, and the actual consumption of the liquid correlated with his mands. The escape response decreased from 35% to 10% from the beginning to the end of the intervention. This decrease indicated that the participant was associating manding for a desired item with the receipt of the desired item. To ensure that this was the case and that the participant had not become prompt dependent, the touch prompt was removed completely in the next phase that served as short-term maintenance phase.


The maintenance phase resulted in the participant' s continued use of mands for an average of 95% of the time. This phase also resulted in a much more even distribution of signs and a decrease in the escape responses to an average of 5%, with a range of 10% to 0%.


A comparative analysis of the total independent mands and the times when the participant gave an escape response during each session was also conducted (see Figure 2). The events in which the participant gave an escape response were grouped into one category entitled "escape response." During the study, the total independent mands occurred at a mid- to high level, with an ascending trend with a mean of 66.04% and a range of 22% to 88.46%, and escape responses were occurring at a mid- to low level with a descending trend with a mean of 24.87% and a range of 10.87% to 58.62%. However, during the maintenance phase, the independent mands peeked at 100%, and the escape responses dropped to an all-time low of 0%. The one-month follow-up yielded promising results, since the data were similar in level and trend to those of the maintenance phase.



The percentage of mands during sessions rose to a higher level with an ascending trend over the course of the training, reaching 100% at one point for independent manding. Frustration was rarely seen as the mands increased, as is evident by the descending trend of escape responses. The design of the study did not allow for instructional prompting, only a verbal prompt to signify the presence of the stimuli because of the nature of the participant's visual impairment and a touch prompt to initiate manding for the desired object. The ascending trend in independent mands during intervention and into the maintenance phase indicates that once the participant was aware of the options that were presented, he was truly manding for them, not asking because he was instructed to "sign eat" or "sign drink." The data also indicate that the participant was pairing the sign "eat" with food and the sign "drink" with soda. The data clearly demonstrate that the participant increased his ability to mand independently. Mand training was a positive intervention for this individual because his mands increased and were maintained while his maladaptive behaviors decreased. As a result, the participant is now able to communicate reliably two of his needs and wants to all his caregivers.

In addition, according to anecdotal reports, there has been a resurgence in previously acquired mands because of the training. Thus, now that the man is consistently pairing his wants with manding, his repertoire is expanding. Future treatment with him will include generalizing the mands he learned in the discrete trial setting to a naturalistic setting through the use of generalization probes outside teaching sessions. Additional teaching sessions will also include strengthening the newly acquired responses and adding more mands to his current repertoire.


Although the study directly taught an individual with autism and vision loss to mand appropriately for desired items, there were some limitations. The limitations included the use of an A-B plus maintenance design. The design was used for ease of replication and the need to teach skills immediately to the participant. However, using a multiple baseline design across mands would have demonstrated more experimental control. Additional limitations included the lack of generalization probes across multiple settings. Future research should conduct generalization probes outside each teaching session and across multiple settings. It should also attempt mand training with other adults and children with severe intellectual disabilities and visual impairments to investigate the generalizability to other similar populations.


American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (rev. 4th ed.). Washington, DC: Author.

Bondy, A. S., & Frost, L. (1994). The Picture Exchange Communication System. Focus on Autistic Behavior, 9, 1-19.

Centers for Disease Control and Prevention. (2007). Autism Spectrum Disorders (ASDs): Data and statistics. Retrieved from http://

Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis (2nd ed.). Upper Saddle River, NJ: Pearson/ Merril-Prentice Hall.

Heward, W. L. (2009). Exceptional children (9th ed.). Columbus, OH: Merrill.

Rosales, R., & Rehfeldt, R. (2007). Contriving transitive conditioned establishing operations to establish derived manding skills in adults with severe developmental disabilities. Journal of Applied Behavior Analysis, 40, 105-121.

Skinner, B. F. (1957). Verbal behavior. Englewood Cliffs, N J: Prentice Hall.

Sundberg, M. L., & Partington, J. W. (1998). Teaching language to children with autism or other developmental disabilities. Danville, CA: Behavior Analysts.

S. Brian Kee, M.S., behavior analyst; mailing address: 6113 West Everett Hill Circle, Arlington, TN 38002; e-mail: . Laura Baylot Casey, Ph.D., assistant professor, Department of Instruction and Leadership, University of Memphis, 404 Ball Hall, Memphis, TN 38152; e-mail: . Clayton R. Cea, M.S., clinical director, Behavioral Services of Tennessee, 1155 Cully Road, Cordova, TN 38018; e-mail: . David F. Bicard, Ph.D., assistant professor, Department of Instruction and Leadership, University of Memphis; e-mail: . Sara E. Bicard, Ph.D., assistant professor, Department of Instruction and Leadership, University of Memphis; e-mail: . Address all correspondence to Dr. Casey.
Gale Copyright: Copyright 2012 Gale, Cengage Learning. All rights reserved.