Matching and Sorting

Matching and Sorting

The matching component of the Matching and Sorting Program outlines steps necessary for teaching the student to match objects, colours, shapes, and behaviours (actions) in three-dimensional (i.e., 3-D object) and two-dimensional (i.e., 2-D picture) combinations (i.e., 3-D to 3-D, 2-D to 2-D, 3-D to 2-D, and 2-D to 3-D). The sorting component of this program teaches the student that some objects group together because they share common properties (e.g., colour or size) or have the same function (e.g., utensils or vehicles). Sorting may be considered an early step in establishing abstract concepts.

The Matching and Sorting Program typically involves the presentation of one or more items displayed on the table. These items are referred to as the sample. The student is handed one item, called the match, which is identical or similar to an item on the table. The student is then given the verbal instruction 'Match' and is taught to place the match on or near (e.g., in front of) the sample. Before the steps for teaching match-to-sample are presented in detail, it may be helpful to note certain advantages inherent in the program. (Note that in the technical literature, what we refer to as the match is often referred to as the sample stimulus, and the sample is often labelled the comparison stimulus.)

We noted in the preceding chapter that one advantage of exposing students to the matching of stimuli is that matching often acquires reinforcing properties for the student (i.e., matching comes to constitute a rein-forcer). Another major advantage of matching is that it helps the student attend to (discriminate) a larger range of stimuli in his environment. Individuals with developmental delays, particularly those diagnosed with autism, are said to be delayed in the skill of attending, often behaving as if they cannot see or hear. Obviously these individuals learn to attend to some stimuli, such as food, puzzle pieces, twirling objects, or floating pieces of lint. However, most stimuli seem to pass them by, as may be the case with typical infants or very young persons. The philosopher William James proposed that the visual world of the infant is a 'buzzing confusion.' Similarly, many professionals consider that typical children at birth do not recognize their parents or 'see' everyday objects. Perhaps, as is implied by these suggestions, all individuals have to learn to attend to their environment.

The task of directing and building attention is one of the most difficult problems adults face in teaching students with developmental delays. Even though a student may look directly at the objects you want him to respond to, there is no guarantee the student truly sees them. It is possible for a student with autism to look directly at a person's face without actually seeing that person, but rather fixating on some minute feature of the face, such as blinking eyelids. As mentioned in Chapter 9, visual orientation is not the same as seeing. Teaching a student to make eye contact with the teacher does not imply that the student is learning to see the teacher. Teaching a student to see (or hear, for that matter) is a slow process. As we understand this problem, seeing and hearing are acquired through discrimination learning. For a detailed description of discrimination learning, see Chapter 16. The way discrimination learning relates to the present program is more fully described in later portions of this chapter. Through the Matching and Sorting Program and the discrimination learning therein, one can make a meaningful step toward teaching the student to attend to visual stimuli. The student will learn to identify and label much of these stimuli in subsequent programs.

The student may make many important gains after learning to match and sort. One of these gains is learning to detect similarities between present events and future ones such that learning can be transferred (generalized) from one situation to another; the detection of similarities among diverse situations helps provide regularity in an individual's behaviour. This gain is significant given that many students with autism and other developmental delays experience difficulty generalizing from one set of stimuli or learning environment to another. Mastery of the Matching and Sorting Program should help reduce such difficulties.

In addition to being useful, the matching or sorting of stimuli is relatively easy to teach. Further, most students enjoy the program and make rapid progress, perhaps because the stimuli are visual or because perceived similarities among objects or events are reinforcing for many students. As may be expected, however, one will encounter large individual differences when teaching students to sort based on abstract properties such as clothing, furniture, foods, or animals. These differences require the teacher to remain flexible when teaching.

Certain prerequisite skills should be established before beginning the Matching and Sorting Program. For one, the student should be able to sit at the table and chair for a few minutes at a time without engaging in too many disruptive behaviours. If matching is reinforcing for the student, he will sit for a longer period of time. It is also helpful for the student to be able to visually orient to the teacher and follow the basic instructions taught in the first hours of treatment (see Chapter 9). In addition, the teacher should be familiar with Chapter 16 and refer to it when and if the student encounters difficulties.

The Matching and Sorting Program can be taught concurrently with such programs as Nonverbal Imitation (Chapter 13), Early Play Skills (Chapter 19), and Reading and Writing (Chapter 29). In the present program, the matching component should be taught before the sorting component. If sorting objects on the basis of common properties is difficult for the student in the early stages of treatment, put this component on hold and reintroduce it after a delay of a few weeks or months. Interim learning about objects may facilitate the acquisition of sorting.

Matching Identical 3-D Objects

Begin the Matching and Sorting Program by teaching the student to match identical 3-D stimuli. When selecting materials for this program, choose from common household objects. We suggest using items the student is likely to have regular contact with, thereby making the skills learned more meaningful. For example, plates, silverware, and small articles of clothing such as gloves may be ideal to use in the early matching tasks. Objects that exist in identical pairs, such as shoes and socks, are also helpful in the early matching programs. In addition, it may be helpful to initially use objects that naturally nest into one another (e.g., cups, plates, spoons). The use of nesting items will remove the ambiguity of where to place the to-be-matched item. Be aware that certain familiar and naturally reinforcing items (e.g., favourite toy figurines) are often difficult to use because the student may want to hold onto and play with them rather than match them to the corresponding objects on the table. If this does not occur, use them.

The steps used to teach matching conform to those specified in the chapter on discrimination learning (Chapter 16). Become familiar with discrimination learning procedures because a thorough understanding of these procedures will facilitate teaching of the skills presented in the current program and every other program in this manual. The present chapter describes how discrimination learning procedures apply to matching and sorting only.

As detailed below, the student is taught SD1-R1 (e.g., matching a cup to a cup) in Step 1 (see Chapter 10 for discussion of discriminative stimulus [SD] and response [R]). In Step 2 the student is taught SD2-R2 (e.g., matching a sock to a sock). Once the student masters both SD1-R1 and SD2-R2, Step 3 is introduced in which SD1 and SD2 are subjected to differential reinforcement and random rotation to ensure that the student attends to the objects presented to her and not inadvertent cues. SD3-R3 (e.g., matching a spoon to a spoon) is then taught. Finally, the student is taught to discriminate SD3 from SD1 and SD2. Make certain that SD1 looks distinctly different from SD2, and that SD3 differs maximally from SD1 and SD2.

During the following steps, the student should sit facing the table and you should sit next to the student. Sitting next to the student makes it easy for you to manually prompt the student. Gradually, as the student requires less assistance and prompting, you may move to sitting across from the student at the table.

To facilitate the following description of the teaching steps, the letters A, B, and C represent the sample (i.e., the items on the table), and A', B', and C represent the match (i.e., the corresponding items that you hand to the student for her to match). Your verbal instruction for this program is 'Match.' Do not include in the instruction the name of the object to be matched (e.g., 'Match cup'). The use of the object label is irrelevant in the beginning stages and may even detract from the student's understanding of the expected response. Initially, be sure to keep the sessions short; the student should sit at the table and chair for only 1 to 3 minutes at a time in the early stages of the program.

► Step 1

Place one cup (Object A) on the table in front of the student. Present SD1, which involves placing an identical cup (Object A') in the student's hand and stating, 'Match,' in a loud and clear voice. As the SD is given, manually prompt the student to make the correct response, placing the cup inside the cup on the table. Reinforce the correct response.

Between every trial (i.e., after the reinforcement of one trial and before the SD of the next trial), remove the stimuli from the table. In this case, remove both cups after reinforcement is given to the student. Such a procedure allows you to present the stimuli again at the beginning of the next trial, thereby maximizing the discreteness of the stimuli presentation. This discreteness increases the likelihood that the student will attend to the stimuli which, in turn, facilitates her learning. Keep in mind that leaving the stimuli on the table expedites learning for some students because it decreases the inter-trial interval.

Repeat the trial and fade the prompt over subsequent trials by gradually decreasing the amount of manual assistance provided to the student. For instance, gently nudge the student's hand with A' in the direction of A on the table, then merely point to where you want her to place the object (i.e., inside the cup on the table). Gradually fade all prompts, and maximize reinforcement for unprompted trials. Place mastery at 5 out of 5 or 9 out of 10 unprompted correct responses.

Next, place the cup in different positions on the table over successive presentations of SD1. For example, place Object A on the left side of the table on one trial, on the right side of the table on the next trial, and toward the back of the table on another trial. This procedure is done so the student learns to place the cup in the cup rather than at a particular place on the table. If the student fails to respond correctly over these trials, move back a bit and introduce the least amount of prompt necessary to reinstate the correct response. Set mastery at 5 out of 5 or 9 out of 10 unprompted correct responses.

As noted in the introduction to this manual, the criterion for mastery varies across and within programs. In general, a relatively easy discrimination may be considered mastered at 5 out of 5 or 9 out of 10 unprompted correct responses, as suggested above. However, given individual differences in rates of learning, some students experience more difficulty than others and mastery may be more likely assured by placing the criterion at 9 out of 10 or 19 out of 20 unprompted correct responses. As a general rule, the first discrimination (i.e., the discrimination between SD1 and SD2) in any program tends to be more difficult to master than later discriminations. Thus, when teaching subsequent discriminations, you may want to set a less stringent criterion (e.g., 3 consecutive unprompted correct responses). If one proceeds too slowly in teaching, the student may become bored and may tantrum or self-stimulate. In addition, by presenting the same SD several times in a row, the problem arises of teaching the student to perseverate by reinforcing repetition of the same response. In contrast, by proceeding too quickly, the early steps may not become fully mastered and, thus, the student may fail at learning later, more advanced steps. The criteria for mastery presented in this manual are provided as guidelines; as you gain experience teaching the student, these guidelines may be modified to fit the student's individual needs.

► Step 2

Choose a second object (B) that is maximally different in colour and shape from the first object. For example, if the first object was a red cup, do not choose a mug or a red toy car as the second object. Instead, a white sock might be a good choice for the second object. Remove the first object (i.e., the cup) from the table. Place one sock (Object B) on the table in front of the student and place the other sock (Object B') in the student's hand while stating, 'Match,' in a loud and clear voice. As SD2 is given, physically assist the student to complete R2 (i.e., placing B' on top of B). Reinforce the correct response, and then remove the stimuli from the student's view. Repeat the presentation of SD2 and begin fading the prompt. Remember to always present the SD (the object and the instruction 'Match') concurrently with the prompt. Consider SD2-R2 to be mastered once the student correctly responds to 5 out of 5 or 9 out of 10 unprompted trials. After this criterion is met, move Object B to random positions on the table as was done with Object A.

► Step 3

Once SD1R l and SD2R2 are acquired separately, SD1 and SD2 should be intermixed to help the student discriminate between them. Begin by placing the cup and the sock on the table with both objects equidistant from the student's midline to minimize position prompts. The stimuli should be clearly visible to the student. Allow an 8- to 10-inch space between the two sets of objects. If the student encounters difficulty, a position prompt may be used initially such that the target object is presented closer to the student than the other object. This position prompt should be gradually faded until the student is able to discriminate between the objects when they are presented at the same distance from the student.

We recommend starting discrimination training with SD1 and prompting the correct response (manually, by position, or both) so as to avoid a non-reinforced trial and the likelihood of a tantrum. An alternative would be to present SD2 first because this association was reinforced last; however, this presentation of SD2 technically would not require discrimination but rather only add to the repetition of the preceding response and hence be of little instructional value.

Intermix SD1 and SD2 according to discrimination learning procedures. Because the student mastered SD2 most recently, it is likely that the student will place the cup on the sock when SD1 (the cup) is presented. To avoid a non-reinforced trial, physically prompt and reinforce the child for Rl (placement of the cup) when you present SD1 (the cup). In addition, to avoid having the student inadvertently learn to place the object in a particular position on the table, randomly interchange the left-right positions of the cup and the sock. If a position prompt is used, randomize the leftright placement of the target object while keeping it closer to the student than the other object. Fade the position prompt over trials by gradually bringing the two objects closer together on each successive trial after fading the physical prompt by providing less and less manual assistance over trials. If the student makes an error on any given trial, do not reinforce the response, but rather end the trial with an informational 'No.' Then repeat the presentation of the object, prompt the correct response, and reinforce.

While fading the prompt, probe with unprompted trials to help reduce prompt dependency, which is likely to occur when reinforcement is provided for prompted trials. Continued reinforcement for prompted trials will decrease the opportunity for the student to acquire the association between the SD and the correct R. Place mastery at 3 to 4 unprompted correct responses in a row with the position of Objects A (the cup) and B (the sock) interchanged and equidistant from the student. Within 2 seconds of completing mastery of SD1-R1, present SD2 (the sock) and simultaneously prompt the student's correct response. Set mastery at 3 to 4 unprompted correct responses in a row.

Over the next several trials, alternate back and forth between SD1 and SD2. With an increase in alternations between SD1 and SD2, you should require fewer and fewer correct successive responses before shifting SD (e.g., shift after 4 unprompted correct responses in a row, then 3, then 2, then 1). Once the student is able to respond correctly to the SD when they are presented singly (i.e., not in blocks of trials), randomly rotate the order of presentation of the objects (e.g., SD1, SD2, SD2, SD1, SD2, SD1, SD1) to help prevent the student from falling into a pattern of responding (see Chapter 16). Also randomize the positions of the objects on the table to avoid teaching the student to respond to a particular position rather than to the objects themselves.

Over successive intermixed and differentially reinforced trials, the student should make fewer and fewer mistakes as the associations between SD1-R1 and SD2-R2 are strengthened because they are reinforced and mistakes such as SD1-R2 and SD2-R1 are weakened because they are not reinforced. Once the student learns to discriminate between SD1 and SD2, we recommend that these SDs be practiced to criterion by all team members. In addition, generalize the mastery across environments by moving the sessions out of the teaching room to other rooms of the house. Cues inadvertently provided by any one teacher or room are less likely to interfere with the student's learning the correct discrimination when these procedures are employed.

► Step 4

The third target (Objects C and C) should be maximally different from the first two objects. For example, if you used a red cup and a white sock as Objects A and B, respectively, let C be a silver spoon. Retain the original setting, with you and the student sitting next to each other facing the table. Place one spoon (Object C) on the table in front of the student. Place an identical spoon (Object C) in the student's hand and present the verbal SD ('Match') in a loud and clear voice. As the SD is given, physically prompt the student to place the spoon (C) on top of or close to the spoon (C) that is on the table. Reinforce the correct response and then remove the objects from the student's view. Repeat the presentation of SD3 and fade the prompt over subsequent trials, remembering to present the prompt concurrently with the SD. Use the same criterion for mastery for SD3-R3 that was used for SD1-R1 and SD2-R2. Once the criterion is met, move Object C to different positions on the table as was done with Objects A and B.

► Step 5

After SD3R3 is mastered when presented alone, intermix and differentially reinforce SD3-R3 first with SD2-R2 (bring to mastery at 5 out of 5 or 9 out of 10 unprompted correct responses) and then with SD1-R1. Once mastery is achieved, intermix and differentially reinforce all three matches (SD1-R1, SD2-R2, and SD3-R3) in a blocked format as done previously (i.e., switch after 4 unprompted correct responses to each SD, then after 3, then after 2, and finally after 1). Randomly rotate the presentation of the SD (once the student can independently respond correctly to the SDs after single presentations). In addition, randomly rotate the left-right positions of the stimuli on the table to eliminate extraneous cues.

It may be helpful in the beginning to have only two samples on the table at any given time (one sample providing the correct match). Gradually, as the student acquires mastery, increase the difficulty by adding to the number of samples displayed on the table. Teach the student to match other 3-D objects by following the same procedures used to teach the first three objects. Eventually you may end up with a dozen different objects on the table (e.g., cup, sock, spoon, toy car, plastic apple, small dish, glasses, pen, toy, doll, key, plastic flower). Give the student one object at a time to match. You are likely to observe the student actively scanning the objects on the table to find the correct match and, perhaps, with a quick and determined move, complete the match.

Once the student learns to match a dozen objects, the difficult part of the matching program is completed, and it is relatively enjoyable for you and the student to progress to other matching tasks. As the student progresses through the various matching steps, you may discover that fewer food and other extrinsic reinforcements are needed to maintain the student's involvement. As is often the case with typically developing individuals, an individual must first be introduced to a certain task, sometimes against the individual's will. However, after exposure to the task, the individual may find it rewarding. The fact that many individuals come to enjoy such things as music, poetry, or sports only after trying them is a good example of the importance of initial exposure to discovering reinforcing activities. If the individual survives the first days and lessons, then she is on her way. If matching appears to be a positive reinforcer for the student you work with, consider using various formats of matching tasks as reinforcement for correct responses in other programs.

Areas of Difficulty

There are several areas in which the student may initially encounter difficulties with the matching component of the Matching and Sorting Program. For one, the student may not know where to place the to-be-matched objects handed to her. The use of nesting objects helps prevent this problem. However, once non-nesting objects are used, this problem of not knowing where to place the items may again arise. Teach the correct placement (i.e., in front of or directly next to the sample for non-nesting items) of the objects either through prompting or by defining an area of space within which to place the object. The latter may be achieved by having the student place matches onto plates or pieces of paper, each containing the correct sample object.

Other problems may be caused by inadvertent prompting. For example, if the teacher consistently uses one hand to give the student the match when the corresponding item is positioned on the left and the other hand when the corresponding placement is on the right, the student may solve the problem by responding to the hand used by the teacher rather than to the match and sample objects. Other inadvertent prompts may occur if the teacher looks at the sample object on the table while handing the student the match object. Similar inadvertent prompts may take place if the teacher signals (e.g., by smiling or shaking his or her head) correct or incorrect responding before the student completes her response. The teacher should wait until the student commits herself to a response (i.e., she places the object and lets go of it) before providing any reinforcement. If the student is unsure of how to respond, it is likely that the student will look for the teacher's inadvertent signs to help guide her to the correct response rather than attend to the target objects. Members of the treatment team should closely monitor one another for potential inadvertent prompts. If such prompts are employed, the student will not learn the skill intended by the teacher.

Matching Identical 2-D Objects

This section describes steps for teaching the student to match identical 2-D stimuli. For some students, it may be easiest to begin by teaching the student to match 2-D representations of the 3-D objects he learned to match in the first phase of this program. For example, if the student was taught to match identical cups, teach him to match photos or drawings of these same cups. Other 2-D stimuli can be obtained from flashcards depicting objects, behaviours, numbers, or letters. Later on, word cards can be used. If 2-D stimuli are taken from magazines, buy two identical magazines and cut and mount pictures in duplicate on index cards (so they are sturdy).

When using pictures, be aware of certain extraneous stimuli. For example, the student may use the borders around some pictures as cues for matching instead of the objects within the pictures. Therefore, remove any borders from pictures. Cut out the pictures as uniformly as possible because the student may match the shape or the size of the pictures rather than the pictures of the objects themselves. For many students, colour is the most salient attribute of the object by which to make the match. When matching identical pictures of a dog, for example, the student may make his match on the basis of a patch of colour on the dog (ending up matching colour to colour) rather than characteristic features of the dog itself.

Although it is virtually impossible to eliminate all irrelevant stimuli, you can work around their interference by following the procedures detailed for introducing non-identical matching presented later in this chapter. What is likely to happen through these procedures is that irrelevant stimuli become unreliable predictors of reinforcement. Hence, attention to these stimuli is weakened while attention to the essential features gains strength because they are consistently associated with reinforcement. It is prudent to keep in mind, however, that the cues you intend for the student to use in solving problems are not necessarily the cues the student will attend to.

Teaching the student to match 2-D stimuli proceeds in the same manner outlined for teaching the matching of 3-D objects. When matching 2-D stimuli, have the student place the picture to be matched on top of the corresponding picture on the table. Some students take great care in making the placement of the picture as exact as possible, pausing to line up the sides and corners in a most precise manner. Such attention to detail is a good sign. It is also evidence of the reinforcing properties of matching.

Matching Colours and Shapes

The procedures for teaching the matching of colours and shapes are identical to those outlined for teaching 3-D and 2-D matching. The stimuli for teaching colour matching should be cards made out of different coloured construction or cardboard paper. Cut the stimuli cards so that they are 5x5 inches in size. Because the student already mastered the matching of objects and pictures, colour matching may proceed relatively quickly.

Once the student learns to match colour cards, 3-D objects to be matched by colour may be introduced. These objects should be identical in all aspects except colour (e.g., large plastic beads or brand new crayons) to avoid ambiguities in the dimension by which the objects are to be matched. If the student has considerable difficulty learning to match colours and makes no progress over several days, the student may be colouring blind. Before reaching such a conclusion, however, make sure that other factors such as the stimuli used, teaching techniques, reinforcement, and prompting are optimal. This is consistent with the suggestion made earlier in this manual to examine the adequacy of teaching methods before attributing failures to the student.

To teach the student to match shapes, cut out various similarly sized shapes from construction or cardboard paper. It is important that all of the shapes are identical in colour so that inadvertent colour cues are not introduced. When beginning to teach shapes, choose two initial shapes that are dissimilar from one another when contrasted (e.g., a circle and a square). Once squares are contrasted with circles, squares may be contrasted with triangles, triangles with diamonds, circles with ovals, and so on, in graduated steps of difficulty.

Numbers, letters, and words are other examples of shapes that can be introduced at this time. Remember, however, that these forms of matching should be introduced systematically; not all three forms should be introduced and targeted at the same time.

Matching 3-D Objects with Identical 2-D Representations

In this section, the student is taught to match objects to pictures of those respective objects. In other words, the student is taught to identify symbolic representations (pictures) of concrete, real-life objects, a skill that is considered by some to be a higher level cognitive process.

This part of the matching program may be of interest for two reasons. First, many professionals propose that higher level cognitive processes cannot be taught on the basis of the learning paradigm (reinforcement theory) set forth in this book. Second, students with autism, pervasive developmental disorder, and similar delays are considered to be deficient in cognitive processes involving abstract reasoning. We mention this here not only because you will enjoy observing the student learn this task, but also because she will do so in spite of what many professionals have proposed. (Adversity often brings out the best in us.) As you progress through this manual, you will learn that higher level cognitive processes can slowly but surely be taught.

Although matching objects to their symbolic representations is basic to the educational process and an important task for the student to learn, making the connection between a 3-D item and its 2-D counterpart is difficult for some students. To facilitate the acquisition of 3-D to 2-D matching, start with objects the student learned to match in previous formats. Because the teaching steps in this section are identical to those presented earlier in this chapter, they are outlined here in relative brevity.

► Step 1

Place the 2-D item (e.g., a picture of a cup) on the table and hand the student the corresponding 3-D object. Say, 'Match,' while prompting the placement of the 3-D object on top of its 2-D counterpart. Set mastery at 5 out of 5 or 9 out of 10 unprompted correct responses. Remove the stimuli between trials.

► Step 2

Place another 2-D item (e.g., a picture of a sock) on the table, and then give the student the corresponding 3-D object. Follow the same procedures described in the preceding step to bring this match to mastery.

► Step 3

Intermix and differentially reinforce the stimuli used in Steps 1 and 2. Once mastered (5 out of 5 or 9 out of 10), continue adding new stimuli as done in previous sections of this program.

After the student learns to match 3-D objects to their 2-D counterparts, teach her to match 2-D pictures to their 3-D counterparts using the same procedures just described. Generalize across teachers and environments.

Matching Objects by Class

Matching No identical Stimuli of the Same Dimension

In this part of the program, the student is taught to match objects to their no identical counterparts. In other words, the student is taught to identify classes of objects and learns that certain objects have some common property even though they may look different (e.g., a shoe is a shoe whether it is brown, white, a loafer, or a sandal). Technically speaking, the student is taught stimulus generalization (i.e., to generalize a response from a particular example of an object to objects that appear similar to the original).

For this task, gather multiple examples of the same object. For instance, use many different exemplars of cups (plastic, paper, and ceramic cups), socks (in assorted colours, sizes, and textures), spoons (metal and plastic), shoes (brown loafers, white sneakers, black sandals, shoes with Velcro, and shoes with laces), and so on. Initially it is a good idea to use no identical items that are very similar (e.g., a white tennis shoe with Velcro adhesives and a white tennis shoe with laces) instead of no identical items that are very dissimilar (e.g., a white tennis shoe and a red high-heeled shoe). The goal is to be able to gradually decrease the physical similarity between the items the student matches until the student can match very dissimilar objects belonging to the same class of objects. Once the student learns to match no identical 3-D objects, teach him to match nonidentical 2-D stimuli.

Teaching should proceed in the same manner described in earlier match-to-sample procedures.

Matching 3-D Objects to Nonidentical 2-D Representations

In this section, the student is taught to match groups of objects to their no identical symbolic (2-D) representations. All of the objects and pictures used in the two previous tasks are needed for teaching this task. Select one item from one class and one item from a different class of 2-D objects and place these two items on the table (e.g., place a picture of a cup and a picture of a tennis shoe on the table). Hands the student 3-D object belonging to one of the classes of objects and instruct him to 'Match' (e.g., hand the student a mug and have him match it to the picture of the cup on the table). Rotate the position of the pictures on the table so as to avoid teaching the student a particular position for placement (e.g., left, right, centre) rather than a match. Once this initial step is mastered (5 out of 5 or 9 out of 10 unprompted correct responses), hand the student a different 3-D object belonging to the class represented by the other item on the table, and teach the student to match it to the appropriate picture.

In practice, teaching may proceed as follows: Place a picture of a man's shoe and a picture of a metal spoon on the table in front of the student. Next, hand the student a woman's shoe while stating, 'Match.' Prompt the correct response if necessary and reinforce the student. Fade the prompt over subsequent trials, randomly rotating the left-right positions of the pictures on the table. Next, give the student a plastic spoon to place, then a white baby's shoe, then a sandal, then a baby spoon, and so on, making sure that all prompts are faded for each object before introducing a new object. When the student learns to match 3-D to 2-D no identical stimuli, teach him to match 2-D to 3-D no identical stimuli following procedures similar to those just described.

Throughout all matching tasks, make sure that the stimuli do not become ambiguous. For example, if a yellow cup and a brown shoe are on the table and the student is given a brown cup to match, he might think that he is supposed to match colours (a salient characteristic of the stimuli) and may therefore match the brown cup to the brown shoe. This would be an incorrect response if the student was supposed to match no identical objects, yet it would be a correct response if he was supposed to match colours. Therefore, make sure the items selected in the early stages of learning are clear in the characteristic by which they are to be matched.

Category Matching

A more advanced aspect of the matching program than matching on the basis of identical or near-identical appearances consists of matching stimuli into groups on the basis of some attribute other than identical or near-identical appearances. That is, objects can be matched on the basis of some common and abstract function, such as things you eat (food), items you put on your body (clothes), or objects that facilitate travel (vehicles). An enormous number of such groupings exist, given that groups can be formed on the basis of one or more common features or attributes. Be aware that learning to match members into groups defined by a common function is difficult and may require some prior conceptual (language) skills. Therefore, the next two sections may be optimally started after the student makes some progress in language.

In this section, teach the matching of commonly occurring categories of objects, such as foods, clothing, vehicles, people, furniture, and animals. It may help to begin with members of categories that differ only slightly from one another. For example, in teaching the student to match the category of animals, start with horses, cows, and dogs, as opposed to chickens, pigs, and whales. Begin with very similar items, gradually extending the concepts you teach to include many heterogeneous members forming one category. Also in the early stages, the categories contrasted should differ maximally from one another. For example, initially refrain from matching fruits versus vegetables.

In illustrating the following steps, the category of animals is taught first and clothing is taught second. Throughout each step, the left-right positions of the stimuli sets should be randomly shifted to help avoid inadvertent position cues.

► Step 1

Place one example of an animal (e.g., a horse) on the table in front of the student. Present SD1, which involves placing another item of the same category (e.g., a cow) in the student's hand and stating, 'Match,' in a loud and clear voice. As the SD is given, prompt the student to make the correct response (placing the cow next to the horse that is on the table). Reinforce the correct response. Remove the stimuli between trials, repeating the trials while fading the prompt. Provide the most amount of reinforcement for unprompted trials. Place mastery at 5 out of 5 or 9 out of 10 unprompted correct responses.

Next, hand the student another example of the animal category (e.g., a goat). Prompt the student to place the goat next to the horse on the table and then reinforce the correct response. Repeat the trial, fading all prompts over the next few trials. Place mastery at 5 out of 5 or 9 out of 10 unprompted correct responses. Continue to teach, one at a time, up to five different exemplars of the same category by following the procedures used to teach the first two exemplars.

► Step 2

Place one example of clothing (e.g., small pants) on the table in front of the student. Place SD2, another clothing item (e.g., a small shirt), in the student's hand and state, 'Match,' in a loud and clear voice. As SD2 is given, physically prompt the student to complete R2 (i.e., placing the shirt on top of the pants on the table). Reinforce the correct response, removing the stimuli between each trial. Repeat the presentation of SD2 and begin fading the prompt. Use the same criterion for mastery for SD2-R2 as was used for SD1-R1. Teach up to five other exemplars of the clothing category, one at a time, using the procedures described in the previous step.

► Step 3

Once SD1R l and SD2R2 are acquired separately, you must help the student discriminate between the two sets of stimuli by adhering to discrimination learning procedures. Place the horse and pants on the table with both objects equidistant from the student's midline and clearly visible to the student. Allow an 8- to 12-inch space between the two sets of objects. Intermix SD1 and SD2 first by presenting repeated trials of SD1 (matching items of the animal category) while fading prompts. Place mastery at 3 to 4 unprompted correct responses in a row. Within seconds after completing mastery of SD1-R1, present SD2 (matching items of the clothing category) and simultaneously prompt the student's correct response. Set mastery at 3 to 4 unprompted correct responses in a row. Over the next several trials, alternate back and forth between blocks of SD1 and SD2 in the same manner, gradually decreasing the number of correct responses in a row required to reach mastery (e.g., 3 unprompted correct responses in a row, then 2, and then 1). Once the student independently responds correctly to each SD when presented directly after the alternate SD, randomly rotate their presentations.

After the student learns to discriminate between SD1 and SD2, we recommend that the SDs be practiced to the criterion for mastery by all team members. Teach other categories using the same procedures described in this section. As always, generalize across teachers and environments.

Sorting

In this section, the student is provided with multiple objects and taught to sort these objects, one at a time, first into groups of identical objects, then into classes based on some common attribute, and finally into categories. This component of the program is similar to those components described earlier, but it is slightly more difficult because the student is confronted with several objects at a time.

Teach the student to select stimuli (objects or pictures) from items presented to her and to place these items into a certain number of containers (e.g., plates) corresponding to the number of sets contained in the stimuli. The initial sorting is based on identical appearances. Begin by finding multiple identical exemplars of objects (e.g., several identical blue blocks and several identical blue socks). Place an example of one object in one container and an example of the other object in another container. The containers should be placed 8 to 12 inches apart and equidistant from the student's midline. Place the remaining blocks and socks together in a pile on the table (no more than three of each item in the beginning) and give the SD ('Sort'). Immediately physically prompt the student to place each of the items into the correct container. Reinforce the correct response, repeat the trial, and begin fading the prompt. Place mastery at twice placing the six items in their respective containers without assistance. Gradually increase the number of objects handed to the student to 10. If the student sorts an item into an incorrect container make sure to deliver an informational 'No' at the moment the response occurs so it is clear to the student that she sorted that item incorrectly. Do not wait until the student sorts all the items before providing a consequence.

After the student masters sorting identical 3-D objects, replace these objects with identical 2-D stimuli. To teach identical 2-D sorting, follow the same procedures used for teaching the student to sort 3-D stimuli. Next, after the student learns to sort identical stimuli, use the stimuli from no identical matching of objects and pictures to teach the student to sort by class. Finally, the stimuli used for category matching should be reintroduced to teach the student to sort stimuli by category.

Concluding Comments

Matching is a powerful educational tool as there is virtually no end to the kinds of concepts one can begin to teach using the matching procedures outlined in this chapter. The Matching and Sorting Program can be used, for example, to teach the student to discriminate among various emotions and behaviours. That is, the student may be taught to match pictures of expressions of different feelings (e.g., happy, sad, angry faces) or activities (e.g., eating, sleeping, driving). Similarly, the student may be taught numerical values by matching cards displaying particular numerals with cards depicting quantities of dots equivalent to those numerals. The dots can later be replaced by objects. Note that certain matching tasks can be very subtle, requiring considerable intellect on the part of the student. Thus, we recommend that parts of this program be maintained and expanded upon throughout the teaching of the remaining programs presented in this manual.