ASA's 36th National Conference on Autism Spectrum Disorders (July 13-16, 2005)

• Recorded Presentation

1. Will verbal requests presented in conjunction with picture communication symbols result in a higher number of correct responses for following directions than verbal requests alone? 2. Will the commands achieved generalize to a novel therapist under those same conditions? 3. Will the commands achieved maintain over time?

The findings of the study contribute to best practice and advances in the field of autism as they examine the conditions under which PCS are effective for facilitating receptive language skills. As the language processing skills of young children with autism are often severely impaired, tasks involving the auditory processing of verbal information typically yield poor performance. The research to date has established that visual systems have improved the expressive language skills of children with autism, because they provide non-transient cues and an alternative means of communication. Research has not, for the most part, assessed the effects of visual systems on the receptive skills of young children with autism.

Methodology Subjects. This study consisted of five subjects each previously diagnosed with autism, ranging in age from five to seven years. All subjects demonstrated significantly delayed skills for their chronological age in following auditory commands, making at least 45 errors on a clinician-made assessment consisting of 128 commands. The commands were presented verbally by the examiner, using no gestures, prompts, or cues. Auditory commands for the purposes of this study were defined as verbal directions presented by an examiner involving body movements (e.g., “stand up”), familiar materials (e.g., “get the ball”), a combination of actions and objects (e.g., “shut the door”), locative concepts (e.g., “put the block in the box”), descriptive concepts (e.g., “show me the one that is wet”), quantity concepts (e.g., “show me the one that has many”), and linguistic concepts (e.g., “give the spoon to her”).

Setting. The study was conducted in a university speech-language pathology center. Four of the subjects received 30 to 45 minute sessions, three times a week; one of the subjects received 60 minute sessions twice a week. Two clinicians who were second year graduate students in speech-language pathology provided the intervention. The assignment of subject to clinician was randomly determined prior to each therapy week, to control for variability in approach and technique. The environment was kept consistent throughout the semester by using the same therapy room, similarly arranged each session.

Research design. An alternating treatments design (Barlow & Hersen, 1984) was used to assess the effect of picture symbols on each subject's response to verbal commands. The sets of items compared were of equal value and the presentation of each specific intervention was counterbalanced. The alternating treatments were the presentation of verbal commands in combination with an associated picture symbol (Treatment A), and the presentation of verbal commands in isolation with no picture symbols present (Treatment B). The picture symbols used in Treatment A were commercially produced, 4 in. (10 cm) black and white line drawings (Mayer-Johnson, 1994) with text accompanying each drawing. Both conditions consisted of verbal commands involving body movements, manipulation of materials, or identification of pictured items, and were presented in alternation during the same session. Each condition contained an equal instructional set and had a baseline measure of zero, indicating that the skills addressed in treatment were not present prior to intervention. Maintenance probes were conducted throughout the intervention to allow the performance of each subject to be measured over time.

Procedures. Only commands not in the subjects' repertoires were presented in the sessions. The focus of the study was on the first 45-50 commands that resulted in an incorrect, inconsistent, or no response from the subject during the pre-intervention. The examiner repeated the command only once and did not use gestural prompts. Subjects were provided with tangible reinforcers determined through direct observation and parent interview to be successful in eliciting optimal performance. The commands were taught following the same format during each treatment condition, with the sole difference being the associated presentation of a picture support for treatment A. All commands were initially presented with graduated levels of support progressing from (a) a command with a physical model to (b) a command with a gestural prompt to (c) a command with no prompts, gestures, or cues. A command was considered mastered when a subject was able to follow the command for five consecutive trials at the independent level (i.e., no prompts, gestures, or cues) during three consecutive sessions without requiring any repetitions, gestures, or physical prompts. The commands were replaced by new, unknown items upon their mastery.

Generalization and maintenance. Short term maintenance probes were conducted throughout Phase 1 of the study to evaluate the short term retention of the commands acquired during the direct intervention. An assessment of previously acquired commands was conducted at the beginning of every sixth session, presenting the directions under their designated condition at the spontaneous level (i.e., no gestural support), two times each. Commands found to be in error through the maintenance probes were reinstated into the teaching condition, following the previously established protocol. Additionally, a generalization phase, Phase 2, evaluated each subject's skills for following directions under their optimal treatment condition, with a novel therapist. During this phase a new clinician was introduced, training and prompting procedures were eliminated, and the subjects were asked to follow all of the commands achieved in the study. Generalization was assessed in one of the final sessions of the study with all commands presented at least twice. Long term maintenance probes were also conducted to evaluate the skills demonstrated during treatment. Two long term probes occurred in Phase 3, following the generalization phase at approximately 10 and 20 weeks.

Results In summary, the results of the study indicate that there was no therapeutic difference between treatments for the subjects' response to verbal requests presented with a picture communication symbol (Treatment A) during the initial acquisition of verbal commands. However, the generalization of acquired commands to a novel examiner was stronger when the subjects were presented with a picture symbol than when presented with a verbal directive alone (Treatment B). Finally, the most visible difference between the treatment conditions occurred in the long term maintenance sessions. All of the subjects retained a greater number of commands over an extended period of time under Treatment A (verbal commands with pictures) than they did under Treatment B (verbal commands alone).

Phase 1. There was a lack of an overall trend during Phase 1 across subjects. Although Subject 4 acquired a greater number of commands under Treatment A (verbal commands with pictures), the difference between conditions was minimal. Subject 1 achieved a slightly greater number of commands under Treatment B (verbal commands alone). Subjects 2, 3 and 5 achieved the same number of commands under each condition.

Generalization phase. Four of the five subjects generalized a greater number of commands under Treatment A than under Treatment B. Only S3 performed 100% of the commands acquired under Treatment A, the condition with picture supports. Two other subjects, S4 and S5, retained 6 out of 7 (86%) and 5 out of 6 (83%) of their previously acquired commands, respectively. The performance of S1, S3, S4, and S5 suggested that their ability to generalize recently acquired commands was stronger when presented with a picture symbol (Treatment A) than when presented with a verbal directive alone (Treatment B).

Maintenance phase. Two types of maintenance probes were conducted throughout the study, short term maintenance probes and long term maintenance probes. Short term maintenance probes were conducted approximately every sixth session throughout the intervention phase to evaluate the retention of acquired commands. Long term maintenance probes were conducted at approximately 10 and 20 weeks following the intervention phase to assess the subjects' retention of acquired commands over an extended period of time.

The number of short term probes varied for each subject depending on the rapidity of their command acquisition. Collectively, the subjects retained more commands for 10 of the short term probes under Treatment A (verbal commands with pictures) compared to 5 of the probes under Treatment B (verbal commands alone). Two long term maintenance probes were conducted during the study. Each subject received two probes with the exception of Subject 5. These long term probes were conducted following a significant lapse in time without direct intervention for command acquisition. Treatment A was most effective for Subjects 2, 3, and 5 as they demonstrated a mean difference of 34.5%, 23.5%, and 33% between conditions, respectively. Treatment A was more effective for Subjects 1 and 4 as well, although the mean difference between conditions was not as pronounced (15.5% and 11.5%).

The results of these long term probes differed from the short term probes as all subjects favored Treatment A The mean percentage of commands retained into long term maintenance under Treatment A ranged from 60% for S1 to 100% for S5. These results were superior to the mean percentage of commands retained into long term maintenance under Treatment B, yielding a 30% to70% range of retention. Overall, Treatment A yielded a mean of 79.3% maintenance of commands acquired in direct intervention when pictures were presented compared to a mean of 55.7% maintenance of commands for Treatment B when verbal commands were presented alone.

Conclusions The subjects in this study benefited from the presentation of visual supports when responding to verbal commands. These results indicated that visual supports, specifically picture communication symbols, are effective in the generalization and maintenance of acquired skills for following verbal directions for young children with autism. The picture symbols did not, however, significantly affect the acquisition of verbal commands in either quantity or efficiency of learning. While the presentation of pictures did not facilitate more rapid skill acquisition for novel commands, once commands were learned, most of the subjects in this study were able to effectively generalize their acquired skills to a novel clinician more efficiently when given the visual prompt of the picture symbols.

Additionally, issues of saliency and iconicity were revealed to be necessary considerations in the development of an augmentative language system. Visual supports other than pictures, as well as tactile input, were noted to facilitate the acquisition of commands, as the subjects benefited from physical manipulation and prompting and the presentation of physical models and gestures. These forms of support not only allowed subjects to learn novel commands, they appeared to also reduce or eliminate the effectiveness of the picture symbols as the physical movements were higher in saliency than the PCS. The iconicity of individual pictures was also found to be a consideration when developing a system of picture support. The subjects in this study had more difficulty achieving commands related to more abstract pictures regardless of the developmental level of difficulty of the specific commands. Overall, the study found that pictures are effective for the generalization and maintenance of verbal commands and that the saliency, iconicity, and frequency of supportive input needs to carefully examined and considered in order for optimal functioning to occur.