| |
Speech disorders in children include articulation and phonological disorders, stuttering, cluttering, developmental apraxia of speech, and a variety of disorders associated with organic conditions such as brain injury (including cerebral palsy), cleft palate, and genetic syndromes. Despite obvious linguistic influences on the analysis, classification, and theoretical understanding of speech disorders in children, most current treatment methods use behavioral techniques. The effectiveness of behavioral treatment techniques in remediating speech disorders in children has been well documented (Onslow, 1993; Bernthal and Bankson, 1998; Hegde, 1998, 2001; Pena-Brooks and Hegde, 2000). Behavioral techniques that apply to all speech discords—and indeed to most disorders of communication—include positive reinforcement and reinforcement schedules, negative reinforcement, instructions, demonstrations, modeling, shaping, prompting, fading, corrective feedback to reduce undesirable responses, and techniques to promote generalized productions and response maintenance.
A basic procedure in implementing behavioral intervention is establishing the baserates of target behaviors. Baserates, or baselines, are systematically measured values of specified behaviors or skills in the absence of planned intervention. Baserates are the natural rates of response when nothing special (such as modeling or explicit positive reinforcement) is programmed. Baserates help establish a stable and reliable response rate against which the effects of a planned intervention or an experimental treatment can be evaluated. The baserate of any parameter should be determined by at least three measures to establish their stability. For instance, to establish the baserates of stuttering in a child, the clinician should measure stuttering in at least three consecutive speech samples. Baserates also should sample responses adequately. For instance, to establish the baserate of production of a phoneme in a child, 15–20 words, phrases, or sentences that contain the target phoneme should be used. Baserates also may be established for different settings, such as the clinic, classroom, and home. In each setting, multiple measurements would be made.
Positive Reinforcement and Reinforcement Schedules
Positive reinforcement is a powerful method of shaping new behaviors or increasing the frequency of low-frequency but desired behaviors. It is a method of selecting and strengthening an individual's behaviors by arranging for certain consequences to occur immediately follow the behavior (Skinner, 1953, 1969, 1974). In using positive reinforcement, the clinician arranges a behavioral contingency, which is an interdependent relationship between a response made in the context of a stimulus array and the consequence that immediately follows it. Therefore, technically, behavioral contingency is the heart of behavioral treatment.
Positive reinforcers are specific events or objects that, following a behavior, increase the future probability of that behavior. Speech-language pathologists routinely use a variety of positive reinforcers in teaching speech skills to children and adults. Praise is a common positive reinforcer. Other positive reinforces include tokens, given for correct responses, that may be exchanged for small gifts. Biofeedback or computer feedback as to the accuracy of response are other forms of positive reinforcement.
Positive reinforcers are initially offered for every correct response, resulting in a continuous reinforcement schedule. When the new response or skill has somewhat stabilized, the reinforcer may be offered for every nth response, resulting in a fixed ratio (FR) schedule. For instance, a child may receive reinforcer for every fifth correct phoneme production (an FR5). Gradually reducing the number of reinforcers with the use of progressively larger ratios will help maintain a skill taught in clinical settings.
Antecedent Control of Target Behaviors
A standard behavioral method is to carefully set the stage for a skill to be taught in treatment sessions. This technique, known as antecedent control, increases the likelihood of a target response by providing stimuli that evoke it. Stimulus manipulations include a variety of procedures, such as modeling, shaping, prompting, and fading.
Modeling
In modeling, the clinician produces the target response, which is then expected to be followed by at least an attempt to produce the same response by the client. The clinician's behavior is the model, which the client attempts to imitate (response). In most cases, modeling is preceded by instructions to the client on how to produce a response, and demonstrations of target responses. Although instructions and demonstrations are part of behavioral treatment procedures, much formal research has focused on modeling as a special stimulus to help establish a new target response.
A child's initial attempt to imitate a target response modeled by the clinician may be more or less correct; nonetheless, the clinician might wish to reinforce all attempts in the right direction. In gradual steps, the clinician may then require responses that are more like the modeled response. To achieve this final result of an imitated response that matches the modeled stimulus, shaping is often used.
Shaping
Whereas straightforward positive reinforcement is effective in increasing the frequency of a low-frequency response, shaping is necessary to create skills that are absent. Shaping, also known as successive approximations, is a procedure to teach new responses in gradual steps. The entire procedure typically includes instructions, demonstrations, modeling, and positive reinforcement. A crucial aspect of shaping is specifying the individual components of a complex response and teaching the components sequentially in a manner that will result in the final target response. For instance, in teaching the correct production of /s/ to a child who has an articulation disorder, the clinician may identify such simplified components of the response as raising the tongue tip to the alveolar ridge, creating a groove along the tongue tip, approximating the two dental arches, blowing air through the tongue-tip groove, and so forth. The child's production of each component response is positively reinforced and practiced several times. Finally, the components are put together to produce the approximation of /s/. Subsequently, and in progressive steps, better approximations of the modeled sound production are reinforced, resulting in an acceptable form of the target response. To further strengthen a newly learned response, the clinician may use the prompting procedure.
Prompting
The probability of a target response that has just emerged with assistance from the previously described procedures may fluctuate from moment to moment. The child may appear unsure and the response rate may be inconsistent. In such cases, prompting will help stabilize that response and increase its frequency. Prompting is a special cue or a stimulus that will help evoke a response from an unsure client. Such cues take various verbal and nonverbal forms. Examples of verbal prompts include such statements as “What do you say to this picture?” “The word starts with a /p/” (both prompt a correct naming or articulatory response). Nonverbal prompts include a variety of facial and hand gestures that suggest a particular target response; the meaning of some gestures may first have to be taught to the child. For instance, a clinician might tell a child who stutters that his or her speech should be slowed down when a particular hand gesture is made. In prompting the production of a phone such as /p/, the clinician may press the two lips together, which may lead to the correct production. Eventually, the influence of such special stimuli as prompts is reduced by a technique called fading.
Fading
Fading is a technique used to gradually withdraw a special stimulus, such as models or prompts, while still maintaining the correct response rate. Abrupt withdrawal of a controlling stimulus will result in failure to respond. In fading, a modeled stimulus or prompt may be reduced in various ways. For instance, a modeling such as “Say I see sun” (in training the production of phoneme /s/) may be shortened by the clinician to “Say I see …”; or the vocal intensity of the prompter or modeler may be reduced such that the modeling becomes progressively softer and eventually inaudible, with only articulatory movements (e.g., correct tongue position) being shown.
Corrective Feedback
Various forms of corrective feedback may be provided to reduce the frequency of incorrect responses. Verbal feedback such as “that is not correct,” “that was bumpy speech” (to correct stuttering in young children), “that was too fast,” and so forth is part of all behavioral treatment programs. Additional corrective procedures include token loss for incorrect responses (tokens earned for correct responses and lost for incorrect ones), and time-out, which includes a brief period of no interaction made contingent on incorrect responses. For instance, every stuttering may be followed by a signal to stop talking for 5 seconds.
Generalization and Maintenance
Behavioral techniques to promote generalized production of speech skills in natural environments and maintenance of those skills over time are important in all clinical work. Teaching clients to self-monitor the production of their newly acquired skills and training significant others to prompt and reinforce those skills at home are among the most effective of the generalization and maintenance techniques.
| |
