Early Receptive Language

Early Receptive Language

In the first hours of treatment (Chapter 9), the student was taught the beginnings of receptive language, namely to respond appropriately to the teacher's verbal requests to 'Sit,' 'Come here,' and 'Drop block.' These exercises are referred to as receptive language tasks because the student is taught to receive the teacher's verbal messages and act appropriately in response to those messages. In technical terminology, the student is presented with verbal stimuli and is taught to make the correct nonverbal response to those stimuli.

NB The present chapter is aimed at teaching more elaborate receptive language than that introduced in the beginning hours of treatment. The student will need to acquire thousands of receptive language messages in order to function adequately in society. Some of these messages are relatively easy for the teacher to teach and for the student to learn, such as 'Come here' and 'Sit.' Others are more difficult, such as 'Finish eating before you put on your coat' and 'We are going to Grandma's house.' Most students are able to learn at least some receptive language as it is presented in this chapter. A smaller portion of students master all of the receptive language programs presented in this manual. The teacher will not know how much receptive language a student will master until some time into the program.

Instructions such as 'Sit,' 'Come here,' and 'Drop block' are ideal to introduce first because the student's responses to these instructions can be manually prompted; that is, the teacher can physically guide the student through the desired responses. Several other early responses can be prompted in a similar manner. These include responses to requests such as 'Clap hands,' 'Push car,' 'Drink juice,' 'Hands quiet,' and 'Give me hug.' Procedures for teaching the student to respond to such instructions are presented later in this chapter.

In teaching the student to respond to requests that require complex behaviours, manual prompting becomes cumbersome or impossible. In such instances, modelling the behaviour is a more practical and efficient prompt. A model prompt is a demonstration of the behaviour for the student to imitate. Examples of responses that are easily prompted by modelling range from relatively simple behaviours, such as sticking out tongue, smiling, blowing, and stomping feet, to more elaborate behaviours, such as writing letters, drawing pictures, playing with toys, combing hair, and playing games (see Chapter 13, which describes how to teach the student to imitate the adult's behaviours).

In this chapter, we progress from simple receptive language (e.g., 'Sit,' 'Come here') to the very complex skill of requiring the student to respond to novel instructions consisting of three actions each. If the student begins to learn the first two receptive language instructions during the first hours of treatment, she will likely master three-part instructions 1 year into treatment. It is important, therefore, that concurrent with the teaching of these skills, the teacher balances the student's curriculum by practicing a number of other programs, such as Matching and Sorting, Nonverbal Imitation, Verbal Imitation, Early Play Skills, Arts and Crafts, and Self-Help Skills.

Assessing Receptive Language

The three instructions learned in the first hours of treatment may be easy for the student to tell apart. However, you may have unknowingly helped the student respond correctly by adding visual prompts, such as looking toward the bucket or block when instructing, 'Drop block,' or stretching out his or her arms when stating, 'Come here.' One way to determine whether such inadvertent prompts are used is for you to fixate on the student's face and refrain from gesturing when giving instructions. An additional procedure involves generalizing the student's mastery across other teachers and adults, because others are unlikely to employ the same inadvertent prompts. If the student responds correctly without the presence of inadvertent prompts to the verbal instructions taught in the first hours of treatment, then go on to Step 1 in the next section. If the student responds incorrectly, teach her to respond appropriately to these instructions by fading inadvertent prompts.

There is no reason to expect that a student with developmental delays can differentiate between two instructions without being taught to do so. To test whether the student can tell the difference between two instructions, present an instruction such as 'Sit' while a block and bucket are on the table and, making certain that inadvertent prompts have been removed, observe the student's response. If the student responds in any way other than sitting down, it is likely that she has not learned to discriminate the instruction to sit from the instruction to place the block in the bucket. When the student can correctly respond to two or more instructions when presented randomly and without prompting, it is said that the student is able to discriminate between these stimuli (i.e., the verbal instructions). In layperson's terminology, it is said that the student gives evidence of paying attention to the differences between these instructions.

Although the understanding of discrimination learning procedures is facilitated by the practical experience of working through the Early Receptive Language Program, it is critical that you first become familiar with discrimination learning procedures as they are presented in Chapter 16. Discrimination training contains important procedures used in all programs, language or otherwise. In this chapter, we describe discrimination learning procedures only as they apply to early receptive language.

The First Two Instructions

Instructions for behaviours requiring object manipulation are generally taught first because the objects help provide clear-cut visual cues for the student and because behaviours that involve objects are easy to prompt. To begin teaching, you and the student should sit next to the student's table and face one another. The student's dominant hand should be nearest the table. In illustrating the following steps, SD1 is 'Drop block,' and SD2 is 'Car.'

► Step 1

Follow the procedures described in Chapter 9 to establish the student's correct response Rl (dropping block in bucket) to your instruction SD1 ('Drop block'). Remember to always reinforce correct responses. Place mastery at 5 out of 5 or 9 out of 10 unprompted correct responses.

► Step 2

Remove the bucket and block from the table and place a toy car on the table such that it is within easy reach of the student. Avoid selecting a car that makes noises or flashes lights as such distractions may detract the student's attention from the task at hand. Present SD2 ('Car'), keeping the instruction short and simple. The correct response, R2, is for the student to push the car back and forth on the table. Say the instruction clearly and in a voice slightly louder than normal conversational tone. Note that, to maximize the discriminability between the two instructions, we recommend using the SD 'Car' instead of 'Push car.' This is done because the use of two-word phrases for each instruction might lessen discriminability.

Up to this point, the student has had no prior experience with this instruction; hence he most likely does not understand what you are asking him to do. You should therefore prompt and reinforce the correct response. Concurrently with or immediately after presenting SD2 ('Car'), place your hand over the student's dominant hand and physically guide the student to push the car back and forth three to four times (enough times to make the response discriminable). Reinforce immediately after the prompted correct response. Repeat this procedure approximately five times, making sure you allow a distinct pause of 1 to 3 seconds between trials.

Systematically fade the physical prompt by gradually reducing it in subsequent presentations of SD2. For example, after the student responds correctly with a full physical prompt for 5 successive trials, physically guide the student's hand to the car and pushes his hand forward once, then let go. The student must still push the car back and forth to receive reinforcement. If the student does not give the required response, the prompt may have been faded too quickly. If this occurs, provide a full physical prompt in 3 to 5 more trials, then try to fade the prompt again by gradually reducing the prompt in each subsequent trial. Begin the fading process by physically guiding the student's hand on top of the car, then letting go. If this prompt is successful, guide the student's hand to a short distance above the car on the next trial. On subsequent trials, lift the student's hand until it is halfway between his lap and the car, and then lift it several inches from his lap, and so on until no prompt is provided at all. If the student cannot follow the instruction independently, return to using the least amount of prompt necessary to reinstate the correct response, and then fade it systematically once more. For students who have made progress in nonverbal imitation, you may prefer to use a model prompt, which is less intrusive than a physical prompt and easier to administer.

As described in Chapter 10, it is sometimes helpful to see if the student can respond to the SD without having to go through the entire prompt fading process. Probe by presenting the SD and withholding the prompt for a trial or two after the student is responding consistently with the prompt. Also, recall that in the early stages of teaching a particular task, both prompted and unprompted responses should be reinforced. When gradually fading a prompt, however, differential reinforcement should be employed such that reinforcement for prompted responses is minimized while reinforcement for unprompted responses is maximized.

After the student responds to SD2 at the criterion of 5 out of 5 or 9 out of 10 unprompted correct trials, begin discrimination training between SD1 and SD2 by intermixing, differentially reinforcing, and randomly rotating the SD as described in the next section.

Intermixing SD1 and SD2

Having just finished teaching SD2 ('Car') to mastery, if you were now to give SD1 ('Drop block') with the car and the block and bucket displayed on the table, the student would most likely in-correctly respond by pushing the car. Such a response would signify that the student cannot yet distinguish the difference between the two instructions and that he will simply perform the last behaviour that received reinforcement. This is a reasonable mistake given that the student has been reinforced for pushing the car over several preceding trials. At this point, all the student hears you say to him is noise; the student understands that you want him to perform an action after you say something. Your task now is to teach the student that the two instructions 'Car' and 'Drop block' have different meanings. In other words, you must now teach the student to discriminate between the two instructions. You can be confident that the student has learned to discriminate between the two instructions when the student can correctly respond to these instructions presented in a random order. The following procedure outlines the steps that lead to such a discrimination.

► Step 3

Place the materials required for responses, dropping and pushing car, on the table. Present SD1 ('Drop block') in mass trials (i.e., provide multiple trials using the same SD). Because the student was most recently reinforced for pushing the car, it is likely that he will make a mistake and push the car when instructed 'Drop block.' If this mistake occurs, the student misses out on a reinforcer. To avoid this, prompt the correct response as soon as SD1 is given. The prompt used should be just intrusive enough to occasion the correct response. In other words, use a partial prompt rather than a full prompt if the partial prompt is effective at occasioning the response. Once the student responds correctly to SD1 in at least 2 consecutive prompted trials, fade the prompt as you did in Step 1. Continue teaching until the student responds correctly in 4 consecutive unprompted trials of SD1. Remember to reinforce correct responses.

► Step 4

Present SD2 ('Car') in mass trials. Prompt the student on the first trial so that he gives the correct response. Fade the prompt. After the student responds correctly without a prompt in 4 consecutive trials, go on to Step 5.

► Step 5

Present SD1 ('Drop block') in mass trials. Prompt the correct response on the first trial. Fade the prompt until the student responds correctly in 3 unprompted trials in a row. Reinforce each response.

► Step 6

Present SD2 ('Car'). Prompt the correct response in the first trial and fade the prompt over subsequent trials. Go on to Step 7 after 3 successful trials without prompting. Reinforce as done above.

► Step 7

Reintroduce SD1 ('Drop block'). On the first trial, test for mastery by withholding the prompt altogether. If the student responds correctly, reinforce him amply and present one more trial of SD1, then go on to Step 8 if he again responds correctly. If the student responds incorrectly, prompt the correct response on the next trial. Fade the prompt over subsequent trials until the student responds correctly in 2 consecutive trials without prompting. Reinforce as done above.

► Step 8

Switch to SD2 ('Car'). On the first trial, probe for mastery by withholding the prompt. If the student gives the correct response, reinforce amply and present 1 more trial of SD2, then go on to Step 9 if he again responds correctly. If the student does not respond correctly, prompt the next trial and fade the prompt over subsequent trials. After 2 consecutive unprompted correct responses, go on to Step 9.

► Step 9

Present SD1 ('Drop block'). On the first trial, probe by withholding the prompt. If the student responds correctly, reinforce and then present SD2. If the student responds incorrectly, provide the least intrusive prompt necessary to occasion correct responding. Fade the prompt and switch to SD2 after 1 unprompted correct response. Follow the same procedure for SD2. Continue alternating between SD1 and SD2 until the student is able to respond correctly without prompting when the SD is directly contrasted with one another (i.e., SD1, SD2, SD1, SD 2).

As the SD1-R1 and SD2-R2 relationships are reinforced and the SD1-R2 and SD2-R1 relationships are not reinforced, the SD1-R1 and SD2-R2 relationships become strong enough to successfully compete with the SD1-R2 and SD2R l associations. Over successive alternated trials involving differential reinforcement, the student should make fewer errors.

As mentioned in the discrimination learning chapter (Chapter 16), there is one problem with systematically alternating between instructions, as in presenting SD1, SD2, SD1, SD2, and so on. The problem is that the student may learn that if SD1-R1 was reinforced on the preceding trial, then he should switch his response on the next trial when SD2 is presented. That is, the student may acquire a win-shift strategy. To avoid such an outcome, introduce random rotation as presented in the next section. For a more complete description of random rotation, refer to Chapter 16.

Random Rotation

► Step 10

Present SD1 and SD2 in random order. For example, first ask the student to drop the block two times in a row, then ask the student to push the car once, then drop the block once, and then push the car three times, and so on. It is important to keep changing the sequence and the frequency of the instructions such that the student cannot discern a pattern and then use this pattern rather than the instruction as a basis for responding. For example, if you present the same instruction many times in a row and reinforce the same response, it may seem to you that the student has learned the correct response. Most likely, however, the student is taught to perseverate; that is, the student is taught to repeat the response that was reinforced on the preceding trial. Students are often very adept at figuring out ways to be successful without having to listen to the teacher's instructions. Make certain that your procedures allow the student to use only your instructions (i.e., your words) to cue responding. Continue to present instructions in a random order until the student responds to the criterion of 9 out of 10 or 19 out of 20 unprompted correct responses.

Mastery of the first discrimination (i.e., the discrimination between SD1 and SD2) constitutes a major achievement for both you and the student in any program. Generalization across teachers and environments is critical in any treatment program, including the present one. We therefore recommend that mastery of the first discrimination be generalized across teachers; that is, these instructions should be practiced by all of the student's teachers so as to reduce idiosyncratic effects of a particular person and thereby help eliminate the effects of potential inadvertent prompts or irrelevant cues. In addition to generalizing across teachers, the SD1-SD2 discrimination should be practiced across different settings by gradually extending the teaching into different parts of the house and outside of the house. If the student makes errors at any time during such generalization training, go back to using the least intrusive prompt necessary to re-establish correct responding before proceeding.

The Third Instruction

To maximize the student's success, select stimuli that are as dissimilar as possible in the early stages of teaching. In later stages, introduce instructions that are similar to each other because the student must eventually learn to respond to subtle details in order to understand much of spoken language. 'Clap' is an appropriate third instruction (SD3) to teach because it sounds and looks different from SD1 ('Drop block') and SD2 ('Car'). We use the instruction 'Clap' to illustrate the introduction of SD3, but keep in mind that your student may have greater success with instructions requiring her to manipulate objects.

► Step 1

Present SD3 ('Clap') and mass trial this SD as you did the first two instructions taught. Let R3 be two to five claps so as to make the response reliably recognizable by all teachers and discriminable to the student. If the student previously learned to imitate clapping, you may prompt by clapping concurrent with or immediately following SD3. If the student fails to imitate your action, prompt by physically holding onto the student's hands and moving them together and apart. Reinforce and gradually fade the prompt. To fade a model prompt, start with your hands at your sides and then bring your hands toward each other until they are about 1 inch apart. Further fading can be accomplished by moving your hands to the level of your chest, then waist, and so on until no visual prompt is provided. When fading a model prompt, it is likely that the student will imitate your behaviour, clapping less and less. To prevent this, do not reinforce the student for partial responses; only reinforce complete responses. To fade a physical prompt, gradually provide less and less manual guidance.

After the student can clap to criterion (5 out of 5 or 9 out of 10 unprompted correct responses), begin random rotation as described in Step 2.

► Step 2

Intermix the third instruction with the first two instructions. Begin by providing differential reinforcement and gradually moving the third instruction (SD3) into random rotation with the first instruction (SD1; as done in Steps 3 through 9 of the previous section 'Intermixing SD1 and SD2'). Place mastery at 9 out of 10 or 19 out of 20 consecutive unprompted correct responses when SD3 and SD1 are randomly intermixed. Then proceed to intermix and randomly rotate SD3 with SD2 to the same criterion for mastery. Finally, use differential reinforcement and random rotation to build the discrimination among all three instructions, setting mastery at 9 out of 10 or 19 out of 20 unprompted correct responses. Generalize this discrimination across teachers and environments.

Additional Instructions

After the student masters three instructions, additional instructions may be taught, such as 'Drink juice,' 'Give me hug,' 'Stand up,' and 'Pat tummy' (additional instructions are listed in Table 15.1). The fourth instruction (SD4) should look and sound as dissimilar from the first three instructions as possible. Teach SD4 following the same procedures used to teach SD1 through SD3. That is, present SD4 in mass trials, using a full prompt on the first trial to help prevent errors and maximize the student's success. Continue to teach SD4 in mass trials until prompts are completely faded and the student responds correctly to criterion. Then gradually intermix this new instruction, with previously taught instructions, following the differential reinforcement and random rotation procedures described earlier. Teach additional instructions in the same manner, introducing them first in mass trials and then gradually mixing each instruction with the instructions taught earlier. Next, generalize the instructions across teachers and environments. After mastery of the first little discrimination, the student should acquire subsequent discriminations with increasing ease.

Additional Instructions for the Early Receptive Language Program

In selecting new instructions, make them as functional for the student and as practical for you as possible. 'Open door,' 'Drink juice,' 'Turn on light,' 'Read book,' 'Feed doll,' and 'Do puzzle' are examples of receptive instructions the student may find enjoyable and functional. The student's newly acquired understanding of language should be to his entertainment and benefit, facilitating adjustment to his social and physical environment.

Teaching 'Drop block,' 'Car,' 'Clap,' and other simple instructions may seem trivial unless it is remembered that, by the teaching of these instructions, the student learns to attend to you and suppress self-stimulatory behaviours. Both are critical achievements for the student. At the same time, you gain valuable experience, acquiring certain basic teaching skills essential to effectively teaching more advanced programs.

Areas of Difficulty

Large individual differences exist among students in their rate of mastery of vocal instructions. Some students move quickly and acquire 20 different instructions over a 3-week span of time; others fail to make even minimal progress in learning receptive language, yet they make rapid progress in acquiring expressive language.

Some students demonstrate difficulty acquiring a particular instruction but make significant progress in learning other instructions. If this occurs with the student you work with, place the particularly difficult instruction on hold (i.e., stop practicing it) for a period of time, replacing the difficult instruction with an alternate instruction to test whether the alternate instruction is easier for the student to learn. It is better to temporarily move away from difficult instructions so as to facilitate success rather than frustrate the student by persisting with instructions that are difficult for her.

Some students experience serious difficulty discriminating SD1 from SD2, making little or no progress after 1 week of training. In such a case, intermix SD1 ('Drop block') with an SD that does not have a verbal component (e.g., use an SD from the Matching and Sorting Program or the Nonverbal Imitation Program). When an SD is used in this manner, it is referred to as a contrasting stimulus. This step should be considered a pre-treatment exercise, giving the student further experience discriminating (attending to, telling apart) the teacher's instructions. This procedure is described in detail in Chapter 16. After the student learns to discriminate SD1 from a contrasting stimulus, reintroduce the previously difficult discrimination.

Some students tantrum a great deal when the teacher intermixes stimuli. If the student has been recently reinforced for responding correctly to SD1 and you now present SD2, it is likely that the student will respond with R l rather than R2. Withholding reinforcement for R l (the response that was previously correct but is now incorrect given the new SD) is likely to be its own SDa frustrating stimulus that cues tantrums. To avoid a non-reinforced trial and an ensuing tantrum due to frustration, prompt the correct response (R2). Although the student needs to learn to tolerate frustration, it may be better to reserve lessons in frustration tolerance for a later time. However, although we have suggested that prompts be used to avoid non-reinforced trials and errors in discrimination learning, there are reasons to believe that errorless learning (which prompting procedures can produce) may result in less robust discriminations than discriminations acquired through both successful and unsuccessful trials. Further, as mentioned earlier, the only way to test whether the student can discriminate among instructions is by observing whether she correctly responds at least 90% of the time to randomly intermixed instructions given without prompts. Therefore, do not be afraid to intermittently probe for mastery by withholding prompts.

As the student masters several receptive instructions and you introduce additional instructions and move on to programs in other areas, make certain that all previously mastered instructions are rehearsed. If previously mastered instructions are not practiced, they will be forgotten. One effective way to ensure rehearsal of previously mastered material is to practice mastered items on an intermittent schedule between new items introduced in other programs. By interspersing previously mastered instructions (which are considered to be easy for the student) with novel items, frustration is reduced and success is maximized.

Instructions Requiring the Student to Leave the Chair

After the student masters approximately 15 to 20 instructions at the table, teach him to respond to instructions that require him to leave the chair. Because some time elapses between the instruction and the student's completion of the response in this format of receptive instructions, the student will have to retain (remember) the instruction. Such retention facilitates the acquisition of skills taught in more advanced programs.

Select a first instruction, such as 'Turn on light.' Teach this instruction as you did the earlier instruction, and maximize success by placing the student directly in front of the light switch. Present the instruction first in mass trials, prompting and then fading the prompt until criterion is reached (5 out of 5 or 9 out of 10 unprompted correct responses). Once this step is mastered, gradually increase the distance between the student and the light switch. Prompt when necessary, and fade the prompt. Eventually the student should be given the instruction while seated at the chair placed across the room from the light switch, requiring that he retain the instruction for several seconds in order to successfully complete the response.

Once the first instruction is mastered when presented alone, intermix it with previously mastered instructions such as 'Drop block.' After the student responds to criterion for this new instruction when it is presented randomly with other mastered instructions, introduce another out-of-chair instruction such as 'Knock on door.' Teach the second instruction using the procedures presented earlier. Continue to teach additional instructions that require the student to leave the chair (e.g., 'Get book,' 'Touch bed,' 'Bring me doll') until the student masters approximately 10 out-of-chair responses.

The instruction 'Come here' may be added to an out-of-chair instruction to create a chain of two behaviours. For example, you could instruct the student to turn on the light and return to you. The completion of the first response (i.e., turning on the light) becomes the SD for the second response (i.e., coming to you). This chain is created by providing reinforcement contingent on the completion

of both behaviours and is prompted by giving the instruction 'Come here' after completion of the first part of the chain.

Two-Part Instructions

After the student masters simple instructions, including those that require leaving the chair and then returning, begin chaining additional behaviours. Do not begin this step until the student masters imitating two-action chains as presented in the Nonverbal Imitation Program (Chapter 13).

Start with a simple two-part instruction, such as 'Clap and stand up,' provided the student has mastered each of the actions composing this instruction when presented separately. Break up the instruction ('Clap and stand up') into two components: Present the first part of the instruction ('Clap') and, just as the student completes the appropriate response, present the second part of the instruction ('and stand up'). This technique is referred to as a pause prompt or a time delay prompt. Fade this prompt by presenting the second half of the instruction closer and closer in time to the first half of the instruction until no delay remains. Provide reinforcement only after the student completes both responses. Should the student engage in the first response and fail to engage in the second response, withhold reinforcement and wait approximately 5 seconds for the second response to occur. If this kind of prompt fails, move to a more intrusive prompt, such as repeating the last part of the instruction ('and stand up') or manually prompting the second response. Continue to teach the two-part instruction until the student can perform both actions in the correct order without beginning either action until the entire instruction has been presented.

A common problem is the student's beginning to respond before the teacher finishes the instruction. For example, the student may clap her hands as soon as the teacher says, 'Clap,' not waiting for the teacher to complete the instruction. Or, if the teacher hurries the instruction by quickly saying, 'Clap and stand up,' the student may stand up without first clapping (the last part of the instruction is the most recent stimulus and therefore the portion the student will most likely respond to). If the student begins her actions before you present the entire instruction, hold the student's hands lightly in her lap and then release her hands as soon as the entire instruction is completed. Holding the student's hands should be considered a prompt, which needs to be faded over time. You can also help the student withhold responding until you complete the two-part instruction by using actions that require object manipulation (e.g., 'Push car and drop block'). Using object manipulation tasks as the first action in response chains allows you to control the timing of the student's response by limiting the student's access to the objects until after the instruction is completed.

Once the student masters the first two-part instruction (i.e., the student reaches the criterion of 5 out of 5 or 9 out of 10 unprompted correct responses), teach a second two-part instruction (e.g., 'Stand up and shake head'). Teach the second instruction in the same manner the first was taught, placing mastery at 5 out of 5 or 9 out of 10 unprompted correct responses.

There is reason to believe that the student may merely learn to perform two actions in a particular sequence rather than listen to the order in which the teacher verbally presents the actions. This is a reasonable outcome given that the teacher reinforces instructions containing a particular order of actions. A good rule of thumb is to suspect that if the student can solve the problem without having to attend to your verbalizations, then it is likely that the student is not attending to the verbal instructions. To help overcome this problem, at some stage in the program, you should make the first action presented in a two-part instruction the same action presented in another two-part instruction. For example, if one instruction is 'Clap and stand up,' another (later) instruction may be 'Clap and touch head.' Teach two-part instructions, keeping the first action in the instruction the same and varying the second action (e.g., 'Clap and drink juice'), until the student responds correctly to the first and second actions of novel instructions without prompting.

Next, continue to teach two-part instructions, but vary the first action while holding the second action constant. For example, use 'Stomp feet' as the second action for the next several instructions while varying the first action (e.g., 'Touch nose and stomp feet,' 'Wave and stomp feet'). Teach subsequent instructions following the same procedures used to teach the first instructions, and place mastery at the same criterion.

To help the student attend even more closely to your instructions (acquire even finer discriminations among your speech), vary the order of any mastered behaviour by sometimes presenting it as the first behaviour and sometimes presenting it as the second behaviour in a two-part instruction. Such instructions are often highly difficult for the student, making it necessary to go back to intrusive levels of prompting when beginning this stage. For example, the second behaviour may need to be presented as the student begins to perform the first behaviour, or the student may need to be given a reminder to wait by having her hands lightly held in her lap.

Teach two-part instructions in this format until the student can respond to any novel two-part instruction on the first trial.

Varying the order of the instructional components allows the student's responses to be brought under the control of fine nuances in your instructions. The goal is to teach the student to attend to your verbal instruction rather than memorize the order of or anticipate the instruction. Mastery of two-part instructions that contain varying components provides some reassurance that the student is not merely learning to chain actions but is listening to the specific parts of the instruction and the order in which its components are presented.

The next step is to present two-part instructions that take more time to complete than earlier instructions (e.g., 'Do puzzle and drink juice'). In the beginning, make sure that the second behaviour is simpler (in this case, not as time consuming) than the first behaviour since the student will have to retain the information as she completes the first behaviour. As the student becomes proficient at this skill, the second behaviour can gradually be made to be more time consuming (e.g., 'Do puzzle and put away toys').

Three-Part Instructions

After the student masters two-part instructions, begin teaching three-part instructions. Do not start this portion of the program until the student masters the imitation of three-action chains because you will use modelling to prompt three-part instructions.

Begin teaching three-part instructions as you taught two-part instructions. Present the first two parts of the instruction and, just as the student completes the second action, present the last (third) part of the instruction. The overlap between the response and the presentation of the last part of the instruction should be faded as was done with two-part instructions. It may be necessary to remind the student to wait by holding his hands lightly in his lap until the instruction is completed. This should also be considered a prompt and, as such, should be faded.

Areas of Difficulty

Regardless of the teacher's skill, a variety of problems can be expected. For example, some students may not make appreciable progress in receptive instructions that require action on the part of the student, such as clapping or stomping feet. However, the same students may make progress in mastering receptive identification of objects, numbers, or letters. Other students may master identifying actions before learning to identify objects. Some students encounter problems 'storing' instructions that have multiple components, only to make progress after gains are made in verbal imitation, particularly if the Verbal Imitation Program produces echolalia repetition of the teacher's instructions. In such a case, the echolalia may facilitate storage of the auditory stimulus through the student's rehearsal of the instruction. Such individual differences require that the teacher be considerably flexible, a skill that comes only with experience in teaching.

Other problems are of a less comprehensive nature. For example, the teacher may find that, as additional responses are taught, the student frequently confuses two particular responses. If this occurs with your student, try to reduce the confusion as follows: First, analyze the instructions to determine if they are similar in some way (e.g., 'Cross arms' and 'Raise arms'). If the two instructions are similar, change one of the instructions to create a dissimilar pair (e.g., change 'Raise arms' to 'Raise hands'). As the student gains proficiency in this program, she will learn to attend to finer nuances in your instructions. During this process, instructions should gradually be made more similar until the student is able to discriminate between such instructions as 'Cross arms' and 'Raise arms.' A second solution to this problem is placing one of the responses on hold for a short time and focusing on teaching a different response. This step should be considered if one of the responses is newly acquired and confusion between the two responses is halting the student's progress in learning additional instructions.

Generalizing Receptive Language

There are four ways to facilitate generalization of the student's mastered receptive language. First, practice the instructions learned in the teaching environment, across settings. Begin by giving the student instructions in the hallway next to the treatment room, then in rooms close to the treatment room, and then move to rooms farther away from the original setting. Go on to practice these instructions in less formal environments, such as in the backyard, at the park, and in the grocery store. As the instructions are moved away from the initial setting and into the community, the student must learn to ignore or suppress all new and distracting cues. For example, if the student self-stimulates on certain features of the bathroom or the grocery store, he must now learn, however gradually, to reduce the impact of such distracting stimuli.

Second, teach instructions that have practical applications in new and varied environments. For example, the instructions 'Get your shoes,' 'Put on table,' and 'Throw away' can be used frequently while spending time with the student around the house. Being creative and using instructions such as 'Put in' (the shopping cart) while at the grocery store or 'Wipe mouth' in restaurants helps the student acquire socially appropriate behaviours.

Third, receptive language may be generalized by the introduction of new teachers. Make certain that these new teachers replicate the original teacher in terms of instructions used and reinforcement delivery. In other words, new teachers should be familiar with the teaching procedures presented in this manual. During generalization, there are bound to be different tones of voice and gestures. Thus, be prepared to step in and regain mastery if there is a loss in correct responding. Learning to respond to many persons increases the probability that the student will respond to novel persons rather than only to familiar teachers. Generalizing instructions across novel persons (i.e., other adults and peers who are unfamiliar with behavioural teaching procedures) is an important criterion for mastery.

Finally, the teacher should eventually vary the delivery of the instructions so the student learns to respond to different versions of the instructions. For example, the mastery of one-part instructions such as 'Clap' may be generalized by gradually adding extra words to form the instruction 'Clap hands' and eventually 'Show Mommy clap hands.' Mastered two-part instructions may be generalized in the same manner. Similarly, the teacher can introduce new versions of instructions that do not necessarily include part of the original wording, such as varying 'Sit' to 'Have a seat.' This instruction may be taught by combining the two instructions into 'Sit, have a seat.' Later, the 'Sit' component may be faded by lowering the volume in which it is spoken, changing the stimulus control to 'Have a seat.' In gradual steps, the instructions may be modified to 'Please come have a seat,' 'Take the weight off your shoulders,' 'Park it,' and so forth. If the student fails to respond correctly, introduce each new version as you would a new instruction by first presenting mass trials of the instruction and then mixing it in with other mastered instructions. Vary instructions based on instructions understood by individuals similar in age to the student. It is better to find commonly occurring phrases to use as variations than to provide exercises involving non-natural phrases just for the sake of generalization.