Summary of Basic Treatment Steps

Summary of Basic Treatment Steps

The preceding chapters provide a great deal of important information. For example, Chapter 9 on establishing cooperation and tantrum reduction introduces concepts basic to behavioural intervention such as instructions, prompts, responses, reinforcement, and shaping. Certain processes such as acquisition and extinction are also introduced in previous chapters. These concepts and processes are used in all the teaching programs presented in this manual.

Before going further, we review these processes and concepts in more detail because knowledge of how they work is important in facilitating the student's success. Although the concepts and processes might seem quite abstract at this time, you will acquire firsthand practical knowledge of them as you begin teaching the student, and you will likely feel that they help you work effectively in a number of situations. Although you may feel awkward at the beginning, over time you will gradually feel more at ease.

Note that the concepts and processes introduced previously are the same ones used in subsequent chapters. To help translate these abstract ideas into effective teaching procedures, we sometimes repeat ourselves; we do this to illustrate how the concepts apply to particular programs and situations. Once you acquire a working knowledge of these concepts and processes, you may be able to generate your own teaching programs when necessary.

Learning to teach effectively is like becoming skilful at any other complex task. No one can complete a complicated task without first learning the basics and then following through with extensive practice and consultation; even the most skilful performers seek consultation and advice from others. When one observes a recognized performer, it seems as though all their moves are executed smoothly with grace and efficiency. With practice, you will become more skilful in your abilities, enhancing the interaction between yourself and the student. However, you may never lose all of your apprehension and anxiety given that so much is at stake. Remember that some level of apprehension can be constructive when used to help monitor your own performance and the performance of others while seeking a higher level of skill as a teacher.

It is extremely important to subject yourself to review by your colleagues; therefore, we place great emphasis on working openly within a treatment team. As recommended in Chapter 4, it is useful to call in an outside consultant who is skilful at delivering behavioural treatment. An experienced consultant can help you sharpen your skills and avoid a great deal of misdirected effort.

Seeking advice in a field as difficult as one-on-one behavioural treatment of students with developmental delays is a measure of one's self-confidence; persons who refuse to solicit feedback likely feel incapable of improving. Avoiding evaluation of one's skills does not allow for betterment, and this ultimately affects the student's progress. In art, athletics, and science, even the most skilful and accomplished person must subject himself or herself to feedback; otherwise, his or her performance deteriorates.

One helpful way to receive constructive feedback is to participate in weekly staff meetings (referred to in Chapter 4). During these meetings, each team member should take turns working with the student for at least a 10-minute period. By demonstrating teaching skills in front of other team members, one's strengths and weaknesses can immediately be recognized. Strengths are modelled for everyone and weaknesses are quickly identified so their deleterious effects do not persist or become magnified. In such a closely supervised and public work environment, it is important not to belittle others for mistakes because sooner or later each team member will make errors. Rather, staff meetings should be used to build skills and unity so that each member feels comfortable working with the team and wants to continue teaching the student.

Another good reason for working within a team is that a single individual cannot become expert in all areas of teaching. Some individuals are particularly good at discovering reinforcers; others excel at suggesting prompts. Some may surpass others in their ability to develop specific programs. In short, there is strength in collaboration.

Introduction to Terminology

Throughout this manual, teaching procedures based on the principles that compose operant learning are used. Operant learning evolved out of what was earlier referred to as instrumental conditioning and trial-and-error learning. Applied Behaviour Analysis is a more recent and updated field of inquiry that includes, but is not limited to, the application of operant learning principles.

Two basic concepts together form the basis of the teaching programs presented in this manual: The discrete trial is described in this chapter, and discrimination learning is introduced in Chapter 16. In the remainder of this chapter, the four parts of the discrete trial are discussed in the following order: (1) the teacher's instruction, (2) the student's response to the instruction, (3) the consequence of the student's response, and (4) the prompt and prompt fading. In addition, shaping, a procedure used to teach new behaviours through successive approximations, and chain' ing, a procedure for building complex behaviours, are discussed toward the end of this chapter. Note that these concepts were introduced in Chapter 9 and are referred to repeatedly throughout the rest of this manual. Given that the four components of the discrete trial are interconnected and are needed for the successful shaping and chaining of behaviours, it is essential that you become familiar with this entire chapter before going on to further teach the student.

Components of a Discrete Trial

The Instruction

Each discrete trial begins with an instruction intended to cue the performance of behaviour. The instruction may be verbal, such as 'Sit down' or 'Come here,' or nonverbal, such as when the teacher shows the student a picture for him to label or a person for him to describe. Further, the instruction may vary from the simple (e.g., 'Sit down') to the complex (e.g., 'You may go out and play after you put away your toys'). Regardless of its form, the instructional component of a discrete trial is referred to in the technical literature as the discriminative stimulus (abbreviated SD). The terms instruction and discriminative stimulus (SD) are used interchangeably throughout this manual.

As the student becomes more familiar with the teaching process, he will learn that when an instruction (SD) is presented, a reward (reinforcement) will be provided for responding correctly to the instruction. In technical terms, the SD is a stimulus that is discriminative for reinforcement (a signal for reinforcement) provided the student responds correctly.

The student no doubt has encountered many instructions, both implicit and explicit, throughout his life, but he has too often failed to understand what the instructions mean. It is now necessary to explicitly teach the student to understand and follow instructions. To facilitate the student's early learning, it is best to begin by simplifying each instruction given to the student. This simplification is accomplished in three ways. First, keep your instructions short and succinct. It is important to refrain from complicated instructions, especially at the beginning of treatment. An example of a complicated instruction is 'Come sit with me for awhile.' This instruction is inefficient in the beginning stages of teaching because it contains too much extraneous information. In the beginning of treatment, the SD should contain only the relevant and critical element of the instruction, which in this case is the instruction 'Sit.' Do not assume that the student can attend to and understand (discriminate) the part of a complex instruction that contains the relevant cues.

Second, decide on the exact wording of the instruction and use this same word or phrase consistently. Consistency in instructions minimizes the student's confusion and maximizes learning. Thus, all teachers should agree upon a consistent format for presenting instructions. Variations in instructions (e.g., 'Be seated,' 'Have a seat,' 'Please sit down,' 'Park your rear here') can be taught later.

Third, present the instruction in a loud and confident voice, making it clearly audible to the student and everyone else in the teaching room. Instructions should not sound as if a question is being asked; reserve the questioning intonation for only those instructions that naturally require its use, such as 'What is he doing?' and 'What colour is it?'

The Response

The student's behaviour cued by an SD is called a response (R). There are three important considerations to make when defining the target response. First, it is essential that all team members agree upon and are consistent in their definition of the response required of the student. For example, if dropping one block in a bucket is the citron for correct responding, this response should be consistently reinforced by all teachers. No one should be allowed to introduce his or her personal requirements, as in having the student drop three or four blocks in a bucket. When teaching the student to sit in the chair after he is asked to sit, make certain that all teachers agree upon the length of time the student is required to remain seated before reinforcement is provided. If the response requirements are altered, as in having the student sit for different lengths of time across different teachers, then those requirements should be agreed upon in advance by all team members. No solo performances of teachers are allowed at this early stage because consistency is critical to the student's learning what he has to do to gain reinforcement. Changing from one response requirement to another minimizes the effectiveness of the reinforcer and risks teaching more than one response at a time, a situation that is too complicated for students with developmental delays in the early stages of these programs. Further, having a clear definition of the response you would like the student to make allows for consistent delivery of reinforcement. The student should be reinforced when his response matches the agreed upon definition of correct responding.

Note that it is extremely important to achieve mastery of behaviour before going on to teach the next behaviour within a specific program. That is, if you choose dropping a block in a bucket as the first response in the Early Receptive Language Program (Chapter 15), make sure the student masters this skill before moving on to another response, such as coming to the chair when requested to do so. The only exception to this takes place when the student fails to make progress toward mastery of the response you have chosen to teach. Altering response requirements is intended to reduce the student's frustration over failure and increase learning.

After giving an instruction, do not allow more than 3 seconds for the student to respond. In other words, the time interval between the SD and the student's response should be between 1 and 3 seconds. This is extremely important in maximizing learning given that, in order for the student to make the connection (association) between the instruction and his response, the two events must occur almost simultaneously in his nervous system. The longer the time interval between the SD and the response, the more likely it is that the student will fail to make the correct connection. Any unintended stimuli that intervene after the SD and before the student's target response may be connected or associated with the instruction.

Failing to respond at all within the 1- to 3-second time interval is considered a non-response. If the student fails to respond, immediately repeat the instruction. If the student again fails to respond, repeat the instruction a third time and simultaneously prompt the correct response. (Prompts and prompt fading are described later in this chapter.)

It is inevitable that the student will sometimes engage in more behaviours than the one you are trying to teach. For example, while teaching the student to sit down, he may resist doing so and simultaneously tantrum. In such a situation, you have little or no choice but to reinforce the sitting response, risking concurrent reinforcement of the tantrum. This problem may eventually resolve itself for the following reasons. First, the sitting response is the only one that is systematically reinforced, and it should therefore eventually become strong enough to replace the tantrum. Second, as the student achieves control over re-in forcers by engaging in appropriate behaviours (e.g., sitting), it is likely that inappropriate behaviours (e.g., tantrum and self-stimulatory behaviour) will decrease. If tantrums do not decrease or you would like to hasten their reduction, first reinforce the student for reliably sitting when asked to do so. Once this step is mastered, proceed to the next step, providing reinforcement contingent on sitting and not tantrum.

Some students evidence major emotional outbursts in the beginning of treatment that show no signs of extinction after 2 to 3 hours of teaching. Such a situation requires considerable ingenuity on the part of the teacher in selecting a response that may reduce or sidestep the emotional outbursts. The following two examples illustrate this problem and describe possible solutions.

Billy

After 2 hours of intensive effort on the part of the teacher, Billy continued screaming and violently protesting, making no progress in sitting. A 15-minute play break was called, and the response requirement was changed from Billy sitting to Billy holding his favourite toy (a train) which the teacher handed to him. Billy accepted this, but only if he stood close to his mother. Over the next 2 hours, the response requirement was gradually changed from the teacher handing the train to Billy to the child having to stretch out his arms to retrieve the train from the teacher's hand. Eventually Billy was required to exert more effort by taking the train from the table (where the teacher was sitting). With this gradual change in response requirements, Billy's cooperation increased and other behaviours could be taught more easily.

Gordon

Gordon tantrumed excessively and failed to make any progress sitting in the chair after 3 hours of intensive effort by the teacher. However, when Gordon sat beside his mother on a sofa and a ring stacker was placed next to Gordon, he would stack rings as they were handed to him by the teacher. The ring stacker was then placed on the floor by the teacher, still near Gordon's mom on the sofa. Gordon moved down to the floor and stacked the rings the teacher handed to him. Over several trials, the ring stacker was eventually placed on the table and Gordon willingly sat on the chair and stacked rings. He was later presented with a puzzle he liked and given puzzle pieces by the teacher.

For both Billy and Gordon, it became increasingly easier for the teacher to introduce new tasks after identifying and working with an initial response tailored for each child. Perhaps the key was helping these children succeed in responding to the teacher's initial request. This success generalized to other teaching tasks.

The Consequence

The consequence provided contingent on the student's correct response is called a reinforcing stimulus (SR). A reinforcing stimulus given contingent on certain behaviour changes the likelihood of the occurrence of that particular behaviour. When a positive reinforces such as food or approval is given contingent on a behaviour that behaviour increases in strength, becoming more likely to recur. Behaviour also increases in strength if it results in the removal of something unpleasant, such as escaping from discomfort like tension, pain, fear, or hunger. This type of reinforcers is called a negative reinforces because it reinforces by the removal of a negative state. Both situations are called rewards in everyday language. Chapter 9, which deals with establishing cooperation and reducing tantrums, illustrates how leaving the teaching situation contingent on sitting in the chair can function as a negative reinforces (strengthening sitting), whereas sitting on a parent's lap and being kissed can function as positive reinforces.

When behaviour occurs and reinforcing stimuli are provided contingent on that behaviour, the behaviour is strengthened. When reinforces are withheld, the behaviour is weakened. That is, presenting contingent reinforcers results in acquisition of behaviour; withholding reinforcers results in extinction of behaviour. These two processes are illustrated in Chapter 5 in the discussion of how self-injurious behaviours may increase or decrease depending on whether attention (a positive reinforcer) is presented or withdrawn. The intent of this manual is to guide the teacher in effectively using reinforcers to increase the student's socially appropriate behaviours while weakening socially inappropriate behaviours, hence helping the student overcome developmental delays.

Remember that each student is different; what may be rewarding for one student may not be rewarding for another student. It is therefore necessary to test reinforcers to determine their efficacy for each individual. A kiss may function as a positive reinforcer for one student but as punishment for another. As mentioned in preceding chapters, nothing is more important in behavioural intervention than gaining access to the individual student's reinforcers, thereby gaining instructional control and effectiveness.

Once you identify a reinforcer that the student accepts when it is provided, you have made the most important advance in helping him. Some students are easier to reinforce than others, but we have never failed to find some item that is reinforcing for any particular student. We usually begin by selecting food reinforcers, and test these to see if the student eats them prior to the teaching situation. If a student accepts the foods you offer, chances are that these food items will serve as effective reinforcers.

After effective reinforcers such as food are identified, explore other reinforcers. Parents typically know best what kinds of foods and activities the student likes and does not like. If the student likes to hold a particular object, let the student receive that object contingent on an appropriate response; that is, use it as a reinforcer. Other activities include matching objects, turning on a light switch, listening to a 5-second tape of a favorite song (e.g., 'Its a Small World'). Do not let the reinforcer consume too much time because this can interfere with or shorten the time needed for teaching. For the most part, a 3- to 5-second duration of the reinforcer is optimal.

Always pair effective reinforcers with social reinforcement. For example, smile and praise the student (e.g., by saying 'Good!' or 'Super!') each time you give him more desired reinforcers (e.g., food or leaving a stressful situation). By doing so, you help the student learn to value social praise as a reward if he does not already do so. In the technical literature, praise is called an acquired, conditioned, learned, or secondary reinforcer, whereas reinforcers such as food and escape from stress are referred to as primary reinforcers. It may help to think of secondary reinforcers as social in nature and primary reinforcers as biological in nature.

Sometimes a student does not accept reinforcers in the early stages of treatment. There are three possible reasons for this. First, when food reinforcers are offered in the early stages of treatment, the student may be too upset to eat or drink. Second, the student may not want to give you control over the situation. Third, the student may not like the particular food, drink, toy, or activity used as the reinforcer. If the student does not accept a potential positive reinforcer, it is likely that he will be reinforced by being allowed to escape from the teaching situation contingent on the desired response. As mentioned previously, escape from an undesirable situation is known as a negative reinforcer and can be used to strengthen (i.e., reinforce) the behaviour it immediately follows.

The student may come to find the company of his parents to be a powerful reinforcer once the student is upset by the demands placed on him. Also, once the student accepts one reinforcer, such as going to his parent(s), he is much more likely to accept another reinforcer, such as food or comfort from other persons. You will discover an enormous range of reinforcers that are effective as you get to know the particular student.

Once you identify a reinforcer, it is essential that this reinforcer be given immediately upon the occurrence of the behaviour you are teaching. For example, if the student is to be reinforced for sitting on the chair for 2 seconds, then after the 2 seconds have elapsed, immediately present the reinforcer. The delay between the completion of a response and the delivery of the reinforcer should be minimal (1 second or less) to maximize the effectiveness of the reinforcer. The longer the delay, the more likely it is that you may inadvertently strengthen an intervening response.

The strength of any given reinforcer varies with exposure to and deprivation of that reinforcer. Weak reinforcers can gain value when the student is deprived of those reinforcers; powerful reinforcers generally lose their value as the student is increasingly exposed to them. In technical terms, the latter phenomenon is known as satiation.

To avoid satiation, vary the types of reinforcers used and provide only small amounts of each particular reinforcer. For example, limit the quantity of a particular reinforcer by giving the student one kernel of popcorn for a correct response rather than a handful of popcorn. Likewise, give the student a small sip or taste of his favourite drink, not half a glass. By varying the use of several reinforcers, the student is inherently allowed limited exposure to each particular type of reinforcer used and satiation is limited or avoided. Vary reinforcers by, for example, tickling the student for one correct response, giving him a favourite toy (for a few seconds) on the next correct response, and hugging him for another correct response. Make certain that the stimuli used as rewards (e.g., tickling, a favourite toy, hugging) are reinforcing for the student, not punishing.

Do not monotonously repeat 'Good' after every correct response, even though it is easy to do so. The student will quickly become satiated on this reinforcer, and this may affect his motivation to learn. When giving social rewards such as smiling and saying 'Good,' exaggerate your expressions. Pretend you are an actor on stage in some large Broadway production.

Once the student masters behaviour and shaping it is no longer necessary, spread out reinforcement over trials by only intermittently using the most powerful ones. Continuous reinforcement (one reinforcer for each response) is used to teach a response; intermittent reinforcement is used to help maintain responding.

Another way to avoid satiation of reinforcers is by using tokens (see Chapter 7). Tokens may be given contingent on a correct response and then later exchanged for a desirable toy, snack, or activity.

To facilitate some degree of desire for reinforcers and help reduce or prevent satiation, limit their availability outside of the teaching situation. For example, if popcorn is used as a reinforcer, limit the student's access to this reinforcer by providing him with it only in the teaching situation and using it in this situation sparingly (e.g., give the student one half or one kernel per correct response). If the student is allowed to have a favourite snack any time he wants, this reinforcement may fail to motivate him during teaching sessions.

When the student responds incorrectly, remove all signs of reinforcers (e.g., place your hand with food in it under the table). The loss of positive reinforcers may function to reduce incorrect responses. Such a procedure is technically known as time-out. After the student incorrectly responds, remove the reinforcers, wait 1 to 2 seconds, and then repeat the instruction. If the student fails on the second try, provide corrective feedback (described below) and then present the instruction a third time, simultaneously prompting the student. Reinforce the correct response.

In consequating incorrect responses, you may want to establish the use of an informational 'No' as corrective feedback. After an incorrect response, say, 'No' in a neutral tone and at the same time looks down and prevent access to all reinforcers. Removing reinforcers may serve to associate 'No' with the removal of positive reinforcers and thereby help establish 'No' as corrective feedback to weaken incorrect responses. This technique is the opposite of the technique used to establish verbal praise, such as saying 'Good,' as rewarding by associating the verbal praise with the presentation of a positive reinforcer or the removal of a negative situation.

To help the student learn to differentiate your feedback, do not say 'No' in the same tone of voice used when saying 'Good' as praise, and do not smile when you say 'No.' For some students, the use of 'No' may pose problems because the word may accidentally have acquired positive reinforcing properties in the past; that is, some students may work hard just to hear you say 'No' and watch you get upset. Conversely, some typical students and some with developmental delays react to 'No' with severe emotional upsets. In such cases, the use of 'No' as feedback is not warranted.

The presentation of a consequence marks the end of a single discrete trial, which begins with an instruction (SD), is followed by a response (R), and ends with a consequence (i.e., a reinforcer [SR] for correct responding, or the removal of a positive reinforcer and/or the use of an informational 'No' for incorrect responding). Before beginning another discrete trial, it is important to allow a brief pause. This pause, the intertrial interval, should be (a) long enough that the presentation of the instruction in the next trial is distinct but (b) not so long that the student becomes unfocused, performing alternative behaviours such as tantruming or self-stimulating. Long intertrial intervals invite self-stimulatory and drifting behaviours; thus, whenever possible, keep the intertrial interval between 1 and 3 seconds (especially in the early stages of treatment). Not only should this help to maintain the student's attention, but shorter intertrial intervals increase the number of trials (i.e., learning opportunities) possible in any given period of time.

Remember that most learning is gradual and incremental and requires several trials. One-trial learning is a rare event for most persons, particularly for those with developmental delays prior to progress in the programs presented in this manual.

The Prompt

Prompts are an integral part of facilitating responses the student may not possess in his behavioural repertoire. Without the use of prompts, the teacher may never obtain the opportunity to reinforce and strengthen certain appropriate behaviours. By definition, a prompt is any action made by the teacher that helps the student performs a correct response so that the behaviour can be reinforced and strengthened. As you become more skilful as a teacher, you will discover hundreds of ways to prompt, ensuring that the student will almost always be successful and you will not be deprived of the opportunity to praise and otherwise reward him. Regardless of the type of prompt used to teach a particular skill, the prompt should occur at the same time as or within I second after the presentation of the SD; if it does not occur within this interval, the instruction and the response may not occur closely enough in time to be-come associated.

A wide range of prompts are suitable for use in the programs described in this manual. The following is a brief summary of the types of prompts you may employ in succeeding chapters. This summary is meant to be a short introduction to prompting; more detailed descriptions of how to specifically use prompts within particular programs are given in the chapters that follow.

Physical or Manual Prompting

In physical or manual prompting, the teacher helps the student respond correctly by providing the student with manual assistance. That is, the teacher guides the student's body through the motions of the correct response. For example, in teaching the student to drop a block into a bucket, the teacher manually guides the student's hand to retrieve a block and then moves his hand with the block in it over to the bucket, helping him release his grip from around the block. When teaching the student to sit in a chair, the teacher places the chair behind the student and the student between his or her legs, and then physically moves the student to a sitting position so that the sitting behaviour can be reinforced.

Modelling

Modelling helps occasion correct responding by providing a visual demonstration of the response. To use modelling as a prompt, the student should first be taught to imitate the adults' behaviours and vocalizations (Chapters 13 and 22, respectively). For example, if you want to use modelling to prompt the student to clap his hands upon your request, first teach him to imitate your various behaviours, including hand clapping. Then instruct the student to clap while you model clapping your hands. Similarly, after the student learns to imitate the adults' vocalizations, modelling can also be used to occasion verbal responses. For example, when teaching the student to identify an apple by name, show the student the apple, ask the student, 'What is it?' and prompt the correct response by saying, 'Apple.' The student imitates your statement of the word, and you reinforce that response. In essence, when using a modelling prompt, you show or otherwise demonstrate to the student the correct response as you would in teaching any typical individual.

In the early stages of learning, the teacher models the correct responses. In later stages, the student's peers may act as models, facilitating appropriate toy play, peer play, conversational speech, and observational learning in the classroom. There are many advantages to using modelling as a prompt; most behaviour you want the student to learn are complex and cannot be helped to occur without modelling. Chapters 13 and 22 on teaching nonverbal imitation and verbal imitation, respectively, therefore contain very important programs for both the teacher and the student to master.

Position Prompting

Position prompting is used to assist the student in programs in which a display of items is presented and the student is required to select the correct item from among these items. To help occasion the correct response, the teacher positions the target item closer to the student than the other objects, thereby increasing the likelihood the student will select the correct item. For example, when teaching a student to identify an apple from a display of items (e.g., fork, shoe, cup, apple), the teacher may place the apple close to the student and the other items farther away from the student. When the SD ('Touch apple') is given, the close proximity of the apple may facilitate correct responding.

Non-specific Prompts

Pointing to, gesturing toward, or glancing at the correct item are referred to collectively as non-specific prompts. These prompts are similar to position prompting in that they are used to assist the student in programs in which a display of objects is presented and the student is required to select the correct object.

Recency Prompting

Recency prompting refers to prompting by linking the target response to a previously learned response. This type of prompt facilitates correct responding by presenting trials of certain SD in close succession. In this situation, the student can use his or the teacher's response in the previous trial as a cue for the next trial. For example, if the student has difficulty naming the object 'car,' present the instruction 'Touch car' just prior to presenting the instruction 'What is it?' The student's hearing the teacher express the word car in the instruction 'Touch car' may help prompt his verbal response ('Car') when the teacher asks, 'What is it?'

Reinforcement-Elicited Prompting

Reinforcement-elicited prompting occurs when rein-forcers cue emotional behaviours (as most reinforcers do). A kiss on the cheek used as a reinforcer can tickle and prompt smiling. Once prompted, the smile can be reinforced and strengthened. Food can serve the same function, and for many students the reinforcing properties of matching (see Chapters 11 and 12) prompt expressions of contentment and more rapid mastery of other behaviours that can be reinforced and strengthened.

Fading Prompts

As discussed in Chapter 9, the main advantage of using prompts is that they help the desired behaviour occur, giving you the opportunity to provide reinforcement. If the student is not prompted, the target behaviour may never occur. If this happens, you will not be able to reinforce the student and the student will not progress. The main disadvantage of prompting is that, when provided with a prompt, the student does not perform correctly by herself; that is; the student is not reinforced for responding to your request but rather for responding to the prompt. With time, the student may become reliant on the prompt for correct responding. Technically speaking, you may reinforce prompt dependency.

To avoid prompt dependency, you must fade all prompts, a process termed prompt fading. Prompts can be faded by reducing their intensity (as in reducing the decibel level of a verbal prompt or lessening the amount of manual guidance used as a physical prompt). Problems may occur when reducing the intensity of the prompt because the student, in order to be reinforced, is inadvertently taught to attend to smaller and smaller units of the prompt which is likely to interfere with her attention to your instruction. One way to resolve this problem is to gradually withhold reinforcement for prompted trials while maximizing reinforcement for unprompted trials. By reducing the intensity of the prompt while concurrently reinforcing the student's response to the SD, the prompt becomes a less reliable cue for responding than the instruction. Reinforcement for responding to the SD is maintained, allowing the SD to become the salient and relevant cue and the strength of the association between the SD and the response to increase. In other words, the student's reliance on the prompt is gradually lessened while the student's reliance on the instruction is gradually increased. In the technical literature, this is referred to as a shift in stimulus control: The transfer of control over the response is shifted from a prompt stimulus to the SD.

Another procedure for reducing prompt dependency is probing the necessity of continued use of a prompt by suddenly withholding the prompt altogether. This strategy determines whether the student can correctly respond to an SD without having to go through an entire prompt fading process. To perform this procedure, use a prompt until the student consistently responds correctly with the prompt, then presents the SD and withholds the prompt for a trial or two. Such trials are called probe trials and are used to determine whether the student has already learned the response and does not require further prompting.

Keep in mind that some responses may be more difficult for the student to learn than others. The student may require minimal prompting to learn one response but may need more prompting to learn another response.

Establishing a Prompt Hierarchy

Successful prompt fading is facilitated by the establishment of a prompt hierarchy. In other words, you should determine a sequential order of prompts for a particular task, beginning with the prompt that provides the most intensive (most effective) assistance and ending with the prompt that provides the least intensive assistance. Once this hierarchy is established, it may be used to gradually and systematically reduce the intensity of the prompt until the student can correctly respond on her own (i.e., without the prompt).

For a teacher to declare that the student has learned a particular skill, the student should respond correctly without the use of prompts. If the student previously learned a certain skill but at some later point responds incorrectly in a trial involving that skill, repeat the same instruction once more to give the student another chance to respond correctly. If the student again responds incorrectly, reinstate the response by repeating the instruction and concurrently prompting the correct response on the third trial using the least intrusive prompt necessary to occasion correct responding. The guiding principle is that you should prompt after no more than two incorrect responses (non-responses are considered incorrect responses).

In the early stages of teaching a stimulusresponse relationship, both prompted and unprompted responses should be reinforced. When gradually fading a prompt, however, reinforcement for prompted responses should be minimized while reinforcement for unprompted responses should be maximized. This procedure is termed differential reinforcement. This term also encompasses the process of reinforcing correct trials (whether or not they are prompted, although prompted trials are reinforced to a lesser degree) and not reinforcing incorrect trials. Once a particular task is mastered, reinforcement should be withheld for trials of that task that require prompting. In this manner, the student learns that independently performed appropriate actions earn greater rewards than assisted actions. A major goal is to teach the student to become more independent and self-sufficient.

Once you become skilful at using the components of a discrete trial, the teaching session resembles an ongoing and uninterrupted dynamic interaction between you and the student. Through such an interaction, you learn, by observing the student's behaviours, how to set the optimal pace for presenting instructions, how to use prompts to help the student perform winning behaviours, and how to provide reinforcers that also act as natural prompts for expressions of positive emotions, which in turn are reinforced, allowing for the prompting of new behaviours and more reinforcement. Under ideal conditions, the teaching situation produces a flow of appropriate behaviours for which the student may be praised, kissed, or otherwise rewarded several times per minute. Such a setting benefits the student in many ways, not only because it helps create contentment, but also because it hastens the student's learning.

Illustrations from Everyday Life

We clarify the universality and importance of prompting by illustrating some uses of prompting in teaching typical individuals. In the life of a person, one usually employs the constructs and processes of prompting without any awareness of doing so. Consider the use of prompting in teaching a child to ride a bicycle. At first, a parent provides a maximum-strength prompt by holding onto the child and the bike to ensure that the bike remains steady and the child does not fall off. Every effort the child makes to ride is praised (reinforced) by his parents and other observers. Over time, as the child becomes more competent at balancing himself on the bike, the parent provides less physical assistance by bracing the bike with one hand rather than two, then gradually releasing the remaining hand from the bike while running alongside just in case the child falls. Finally, the child is allowed to ride without assistance.

Many other prompts are used when raising children, many of which are more subtle than physical prompts. For example, if one asks a child what time it is and the child does not respond, one often prompts by pointing to the dial of a clock. If this minimal amount of prompting is not effective, one may resort to a more intrusive prompt by giving the full answer, 'It is 12 o'clock.' If a child asks her mother for a favour (e.g., 'I want ice cream') and the mother prompts, 'Say please,' this prompt may later be faded into a less intrusive prompt such as 'What's the magic word?' Parents may say to their child, 'Please help your sister.' If the child fails to do so, a parent may prompt either by physically guiding the child through the motions or by modelling the desired behaviours. Almost all answers to questions requiring factual information (e.g., 'Who is the president of the United States?' 'What is the capital of France?') are prompted when first taught. If the correct responses were not prompted, the child would not know the answer.

Other prompts are less obvious, as when a mother smiles and asks her young child, 'Do you love Mommy?' and the child responds with a slight echo, 'Love Mommy,' which is reinforced. Another example is stating, 'Come give Daddy a kiss,' which may be a prompt in the sense that this statement can be faded until the mere sight of Daddy occasions a kiss. We call such a kiss spontaneous. Few parents wait around for their children to show love to each other or to themselves as parents. Parents prompt and reinforce such behaviours; otherwise, they may rarely occur and one may raise a child who demonstrates little love.

What has been said about raising children can apply equally well to maintaining a happy marriage. If your partner is somewhat distant or aloof, he or she may rarely show affection. If this occurs, your marriage may not last; it may be placed on extinction or at least it may not be very happy. To remedy this problem, it helps to prompt and reinforce affectionate behaviours. For example, one may model a desired response and comment, 'See, this is how you do it,' or use a less intrusive prompt such as asking, 'Do you really love me?' Examples of more subtle prompts can be observed in arrangements of the physical environment. Sleeping in one bed, or at least in the same room, is an example. If a husband sits on one side of the sofa intently watching football on television and his wife sits on the opposite side of the sofa, the latter may position a pillow close to her and comment, while patting the pillow, 'This is better for your back.' The intent is to prompt closeness without being too intrusive. You may know of many other subtle prompts. Such prompts must eventually be faded, however, or the partner's attention may not appear genuine or natural. Prompts appear everywhere in day-to-day life, facilitating the occurrence of desirable behaviours to be reinforced and strengthened.

Just as prompts occur in everyday life, constructs such as reinforcement can be observed in day-to-day living. For example, without regular reinforcement (e.g., kisses, hugs, compliments), a marriage may not last. How often should one reinforce one's partner? No doubt there are individual differences, varying from once a day to once an hour. What behaviours and attributes should one reinforce? These are easy to discover, once one starts: 'Your hair is beautiful,' 'I like your family,' 'You are smart,' 'Good dinner,' 'You are so strong,' 'I can trust you,' and thousands of other comments. Similarly, without the expression of love to children, they may not learn to love you. In the case of raising typical children, the details of how these constructs operate may not need to be known; typical children become attached to others and develop normally without caregivers knowing much about the processes underlying their learning. Parents of children with developmental delays love these children as much as they do their typically developing children. In the case of the former, however, love is not enough. The operation of learning-based constructs must be known in great detail for a parent or teacher to be of significant help.

Shaping

Shaping is a procedure used to teach new behaviours through a process of reinforcing successive approximations of a desired or target behaviour. In other words, the student initially is reinforced for a behaviour that is close to the target behaviour. After the student consistently performs this approximation, the criterion for receiving reinforcement is gradually changed to the student's performing a behaviour that is more similar to the target behaviour. Over time, the criterion for receiving reinforcement is changed until only the target behaviour is reinforced. The advantage of this procedure is that it allows the student to be reinforced immediately and it helps reduce frustration for the student by initially focusing on easier, more attainable goals. One of its disadvantages, however, is that by implementing this procedure, it may take a long time to teach the target behaviour. The use of shaping as it applies to teaching specific skills is detailed in subsequent chapters.

Chaining

Chaining is the creation of a complex behaviour through the combination of simple behaviours into a sequence that forms a single, more complex behaviour. This sequence of behaviours is called a chain. The term chain is used because earlier units of behaviour come to provide the cues for the subsequent units of behaviour. In technical terms it can be said that the feedback generated by one response (technically known as response-produced stimulus) provides the stimulus for the next response.

A chain is created through a link of behaviours, each link being the neural associations between one response and the next. To chain behaviours, one must follow four steps: (1) identify the target behaviour; (2) break down the target response into small discrete steps; (3) teach the individual steps using a separate instruction and reinforcement for each step; and (4) fade the instruction, prompt, and reinforcement for each individual step until there is only one instruction at the beginning of the chain and one reinforcement at the end of the chain.

Two types of chaining are presented in this manual: forward chaining and backward chaining. Forward chaining entails teaching and combining behaviours in a forward chronological order. Teaching begins with the first step in the chain and ends with the last step. Backward chaining involves teaching and combining behaviours in backward chronological order. The last behaviour in the chain is taught first and the remaining behaviours are added in descending order. Backward chaining requires that in the beginning stages, the student complete only the last step(s) in a chain before he is reinforced, whereas forward chaining works in the opposite direction.

The extent to which a behaviour needs to be chained in a piecemeal fashion and prompted manually depends to some extent on the student's mastery of nonverbal imitation (Chapter 13). For example, the adult might have to model and physically prompt each step for a less advanced student. For students who have acquired nonverbal imitation (Chapter 13), it is possible for the adult to model the entire behavioural sequence and 'hand shape' those elements that need improvement. Chaining techniques are explained in more detail in several programs in this manual.

Using Discrete Trials

During the beginning stages of teaching, keep the teaching sessions to a maximum of five to six trials, taking short play breaks between sets of trials. Sitting in the beginning stages should not exceed 2 minutes between breaks. As the student advances through the programs in this manual, you may extend the time spent in the chair from under 1 minute to up to 5 minutes, allowing for more trials. Eventually, as done in school, the student may be required to sit for 45 to 50 minutes at a stretch. When it is time for a play break, signal the break with the instruction 'All done' and concurrently help the student leave the chair to go and play. In all programs, signal a play break contingent only on a correct response (unprompted whenever possible). Make certain that the student does not receive a break contingent on non-compliance, a tantrum, or an incorrect response.

Purpose of Discrete Trials

The use of discrete trials in teaching students with developmental delays is beneficial for four reasons. First, discrete trials clearly delineate what you are trying to teach and help the student attend to the SD (i.e., the instruction). Second, discrete trials let you and the student know immediately whether the response is correct. Third, discrete trials help you instruct the student in a consistent manner; this comes about because each unit is specified in concrete detail. Consistency is crucial in the beginning stages of teaching because any discrepancies may cause the student to become confused, delaying her progress in learning. Fourth, discrete trials allow for quick and easy assessment of the student's advancement. A very sensitive measure of learning is the rate at which the prompt can be faded across trials. Another measure is the speed with which the student responds correctly to your instructions across new and different tasks. Chapter 33 describes techniques for collecting data to document progress in learning.

Discrete trial procedures are frequently critiqued by those who believe that such procedures fail to teach higher level skills such as language, fail to produce generalization, fail to teach functional skills, inadequately use reinforcers, place stress on parents, and are inefficient and slow moving (Koegel &. Koegel, 1995). By way of rebuttal, it should be noted that discrete trial procedures form the basis for teaching abstract language, including concepts such as colours, prepositions, pronouns, time, and cause-effect relationships, as well as conversational speech and a host of other complex skills. Discrete trials also form the basis for teaching imitation and observational learning. Both of these skills help students learn on their own through attention to the behaviours of other individuals. It should be further noted that most students graduate from reliance on discrete trial procedures as they progress through the programs presented in this manual. Finally, treatments that rely on discrete trials are the only ones that are supported by favourable long-term outcomes as measured on socially and educationally meaningful and well-normed objective assessment instruments.

Some persons claim they are against behaviour modification. The fact is, everyone continually modifies others. The challenge is to learn how and when to use behaviour modification principles in a constructive manner. Many people do not know how to effectively use such principles, especially when it comes to teaching individuals (whether they are typical or developmentally delayed). However, everyone can be taught to do so. The information presented in this manual is aimed at helping you learn constructive application of behavioural intervention techniques.

Concluding Comments

When exposed to behavioural teaching programs, persons with developmental delays evidence a great deal more complexity and individual differences than can be accounted for by current clinical and educational theory or by what is presented in this and similar teaching manuals. Thus, as programs are presented, difficulties will be encountered. Solutions to examples of such problems are reviewed in most chapters as they relate to specific programs and in Chapter 35 on common problems in teaching.

As the student experiences difficulties, you will also be challenged. Thus, it is important to remember that no one becomes a master teacher overnight. One has to make a beginning; mistakes are as inevitable as progress. Progress occurs in a situation where you are open to feedback. Just remember that no errors are so detrimental that they cannot be repaired.