1	TCLC Summer Institute Overview of Techniques, Methods, and Data
2	Individuals Are Treated Individuals are treated not groups. A single approach is not applied to everyone in treatment. Techniques for training are chosen on the basis of effectiveness. All behaviors of relevance should be addressed.
3	What Is Behavior Analysis? Behavior analysis is the science that is devoted to the study of the optimal learning environment. It calls behavior “operant” because it operates on the environment.
4	What Is Behavior Analysis? This science breaks the context into establishing operations, antecedents, the behavior, and its consequences. The Antecedent-Behavior relationship is referred to as stimulus control or in some literature it is referred to as the opportunity to respond.
5	What Is Behavior Analysis? Opportunity to respond is the effect of your instructional antecedent on gaining the desired response. Opportunity to respond tells us much about the importance of instructional matching (discussed more later).
6	What Is Behavior Analysis? The Behavior -Consequence relationship is often referred to as the functional relationship. A complete A-B-C relationship is referred to as the learned unit.
7	Behavior Management
8	Different Popular Techniques DTT - Discrete Trial Training NET – Natural Environment Training DI – Direct Instruction MLT – Milieu Language Training
9	Different Popular Techniques Incidental Teaching NLP – Natural Language Paradigm Free Operant Techniques Shaping Techniques Chaining Techniques
10	Quality in an ABA Program Uses collected data to make treatment decisions. Refines procedures (protocols) to maximize benefit. Does not ignore the “function” of behavior in the natural environment. Inappropriate behavior. Verbal Behavior. Generalization of Learned Behaviors.
11	What Should an Analyst Do? Train individuals to learn from the natural environment. Train individuals to function normally “normalize” in their environment. End treatment as quickly as possible.
12	The Three Term Contingency S^D= Discriminative Stimulus R = Response S^R = Reinforcing Stimulus
13	The Discrete Trial S^D^è R ^èS^R^è^ITI S^D= Discriminative Stimulus R = Response S^R = Reinforcing Stimulus ITI = Inter-trial interval
14	S^D= Discriminative Stimulus Stimulus designed to evoke a particular response. Make the S^D simple, clear, and concise. Say the entire S^Dwithout interruption. Decide on the exact wording and use consistently. Use a consistent format when presenting materials. Do not repeat the S^D("Clap, Clap, Clap") without consequating the child's response. S^D's should initially be louder than your typical speech, then faded to more natural language.
15	R = Response The child’s action in response to an S^D. Define what is expected for the response before you give the S^D. Correct responses. Before beginning a lesson, read the log book for most recent response criterion considered as correct. Be consistent in what you are requiring as a correct response. Response criterion must be consistent across all therapists. As response criterion is shifted to higher levels of independent functioning, document changes in the log book.
16	R = Response Incorrect responses. A response not meeting specified criteria is incorrect. Non-response is an incorrect response. (Give the child 3-5 seconds to respond after presenting the S^D). Multiple responses are considered incorrect.
17	S^R = Reinforcing Stimulus The stimulus following the response. A reinforcing stimulus will increase the likelihood of the response in the future, while an aversive stimulus will decrease the likelihood of the behavior in the future. Follow correct responses with a reward, and incorrect responses with an informational “nope”. The “nope” is simply stimulus designed to indicate that no reinforcement is forthcoming. Reinforcers must be delivered immediately following the response.
18	S^R = Reinforcing Stimulus Avoid satiation by using a variety of rewards, given in small amounts. To avoid confusing the child, consequences for correct and incorrect responses should be easily distinguishable from each other. Common mistakes are: saying “nope” in the same tone of voice as “good” and smiling when saying “nope”. Vary your reinforcers: Tickle, hug, make it “circus time” - don’t repeat the same reinforcer continuously (e.g. “good” with the same inflection).
19	ITI = Inter-trial Interval Stimulus designed to indicate the end of one opportunity to receive reinforcement and the beginning of the next. This is a 1-to-3 second pause between each discrete trial. It should be long enough that each trial is distinct, but not so long that the child loses focus or engages in alternative behaviors. During this time the field is cleared or reset to emphasize the new opportunity.
20	S^P= Stimulus Prompt Stimulus designed to help the child achieve the correct response. All prompts must be faded over time before a response is considered mastered. S^D_(SP)è R èS^R The prompt occurs paired with or immediately following the S^D. Stimulus prompts are used when the child has been incorrect over two trials (No - No - Prompt Sequence). Or used consistently and with a planned fading procedure in an errorless format.
21	Types of Prompts Simplify the S^D by removing all unnecessary information. Over Emphasis / Exaggeration Physical Model Positional / Proximity Instructional / Explanation Verbal modeling (Verbal Prompt) Inflectional / Tonal Gestured Pointing Verbal
22	Prompting Increases likelihood that behavior will occur (helps child perform). Allows child to get more positive reinforcement and thus makes learning an overall reinforcing experience. Allows the child to experience the concept to completion first hand, adding to the effect of making the connection between the stimuli, discrimination and the response. Decreases frustration.
23	Disadvantages Prompting dependency, “hand-holding”, expecting prompt: usually happens when prompt is not faded soon enough. Inadvertent prompting, usually prompts child to be incorrect.
24	Rules for the Use of Prompts Use least intrusive prompt possible to elicit response because: Fades easier. Least intrusive prompt is more like natural environment and will improve generalization.
25	Rules for the Use of Prompts Fade prompt ASAP - so child won’t become prompt-dependent. Use differential reinforcement - so child will work harder to give correct response without prompt. Change prompt to prevent prompt dependence. Avoid inadvertent prompts. Examples: Looking at the correct object, nodding yes or no as the child moves towards the stimuli, or even pointing to the correct objects.
26	Fading The removal of the unnatural cue or prompt that allows the child to respond to a more natural SD. RULES FOR FADING. Fade ASAP ( Ideally no more than five steps to fade the prompt completely). Fade prompts gradually. If you have to repeatedly go back to a heavy prompt, you are fading to fast.
27	Fading Fade from most intrusive to least intrusive prompts. Spend time preparing for the use of prompts including but not limited to: Level of prompting used in assisting the child to emit the correct response. Fading strategy for each prompt used. A strategy for documenting the level of prompting used in each sitting. Separate strategies to vary the delivery of prompts.
28	Training & Methods
29	Baseline Measurements Operant level refers to the frequency of a behavior prior to treatment and requires measurement of a naturally occurring behavior. In order to measure the effectiveness of our intervention we must conduct a baseline. A baseline is a period of time that the behavior is measured prior to beginning any treatment. We can conduct the baseline in a couple of different manners. Frequency counting refers to recording the number of times a behavior occurs during a specified period of time. Time sampling refers to observing the client at fixed intervals of time and recording the occurrence or absence of the behavior during that time.
30	Reinforcement Schedules Schedule of Reinforcement is a rule specifying which occurrences of a given behavior, if any, will be reinforced. The two simplest schedules are Continuous Reinforcement and Extinction. Intermittent Reinforcement refers to the maintenance of a behavior by reinforcing it only occasionally (i.e., intermittently) rather than every time it occurs. An Intermittent-Reinforcement Schedule is any rule specifying a procedure for occasionally reinforcing a behavior.
31	Reinforcement Schedules Intermittent Schedules that increase and maintain behavior have several advantages over continuous reinforcement: Behavior that has been reinforced intermittently persists more readily when transferred to reinforcers in the natural environment. Behavior that has been reinforced intermittently takes longer to extinguish than behavior that has been continuously reinforced. The reinforcer remains effective longer than with continuous reinforcement because satiation takes place more slowly. Individuals work more consistently on certain intermittent schedules than on continuous reinforcement.
32	Verbal Behavior In 1957 B. F. Skinner completed and published his book VERBAL BEHAVIOR and outlined a systematic approach to the development of language in the human species. Skinner provided an analysis of language based on the same environmental principles of behavior he previously used to analyze non-verbal behavior (Skinner, 1938, 1953).
33	Verbal Behavior Skinner makes the theoretical point that language is behavior caused by environmental variables such as reinforcement, motivation, extinction and punishment. This development was not viewed as a “cognitive ability” but, instead as a set of behaviors that contextually ensure that people become members of a verbal community.
34	The Operants Mimetic Mand Echoic Tact Intraverbal Autoclitics
35	The Mimetic: Imitation, this behavior is non-verbal in form. It is the tendency to imitate someone else’s modeled motor movements. The tendency to “clap” when someone else “claps”. Generalized reinforcers / Automatic reinforcement.
36	The Mand: Not all language functions should be brought under SD control. The Mand is a verbal operant that is under the functional control of establishing operations. Mands, regardless of their appearance or form, function as demands, commands, or requests. Perhaps the easiest way to define a mand is that this verbal behavior specifies the reinforcement. The word “water” when one is thirsty is a mand. It is possible of course to condition or contrive these establishing operations in order to teach mands.
37	The Mand: Mands take many forms. People mand for items, attention, information and relief from aversive stimuli for example.
38	Establishing Operations: Establishing Operations (EO) can be defined as an environmental event or operation that affects the individual by momentarily altering the reinforcing effectiveness of other events. EO also affects the frequency of behavior that has been reinforced by those other events. Establishing Operations can be utilized to teach many verbal responses by making a previously neutral stimulus function as a reinforcer. Proper manipulations of EO may evoke behavior by changing what functions as reinforcement.
39	Establishing Operations: Improper applications of the principles of EO can result in verbal responses becoming or remaining under the functional control of SDs. Simply asking "What do you want?" or "What do you need?" at an inappropriate moment can result in the failure to bring verbal responses under EO control.
40	SD or EO? As both Establishing Operations (EO) and Discriminative Stimuli (SDs) evoke behavioral responses, it is often difficult to identify which is at work.
41	SD or EO? The differences between EO and SDs relate mostly to the process of bringing specific responses under functional control of specific stimuli. With an SD this control is established because the stimulus (instruction) is always correlated with a high frequency of reinforcement (rewards).
42	SD or EO? EO does not make reinforcement more available, it makes it more valuable to the individual at a particular time. And this change in value affects the motivational variables that increase the likelihood of an individual emitting a behavioral response at a particular time.
43	SD or EO? Responses under the control of EO occur due to environmental events or operations, not specific instructions. An SD may be required to alter the likelihood that a mand is emitted, but the EO must ultimately control the response.
44	The Echoic: The echoic behavior class is essentially the ability to repeat or “echo” back verbal behavior or patterns of phonetic sounds in the presence of those similar sounds. The typical reinforcement that is received for this type of behavior in children is social reinforcement. It is a basic skill of building verbal behavior and indeed can be developed in children who have never had an echoic through the use of the mimetic, non-contingent reinforcement , automatic reinforcement procedures and even through the use of augmentative communication. It is simply the tendency to say “cat” when you hear someone else say “cat”.
45	The Tact: The tact is a behavior class that includes identifying objects, actions, events, attributes or other observable conditions in the presence of those conditions. Simply it is the tendency to say “water” when one sees “water”. Or “running water” when one sees “running water”.
46	Intraverbal: Essentially here the operant class is under the control of other verbal behavior, this is where conversation occurs. It is simply the tendency to say “water” when someone else says, “drink”. The intraverbal is a behavior class that includes responding that is controlled by non-matching verbal stimuli. Simply it is the tendency to say “lamb” when one hears “Mary had a little _____” or to say “weasel” when one hears “pop goes the ______”. Many forms of IQ testing seem to actually be measures of the intraverbal response class. Intraverbal behavior is the basis or primary behavior in social interaction and conversations between individuals.
47	Verbal Behavior Did you know that Skinner defined the functional relations in his program as relations between people and not individual processes? This is an often-missed fact. For example, a mand is only a mand if the listener responds to it as such.
48	Receptive: Receptive skills are behavior that humans emit as the “listener”. They can include responding to the name, feature, function, class or other quantifying identification element as well as actions on the part of the listener.
49	Receptive: It is essentially the tendency to touch “water” when someone else says “water?”. Zettle and Hayes have completed a parallel analysis of listener behavior to Skinner’s analysis of speaker behavior.
50	Receptive: In Hayes’s analysis pliance is the listener’s response to a mand, tracking is the listeners response to the tact, augmenting the response to the intraverbal. Often times many programs fail to use more comprehensive analysis of listener behavior in programming for children with autism.
51	Receptive: This leads to gaps in children’s response patterns. This failure in programming might lead children to look like ADHD children when the begin to habilatate. Strong receptive understanding benefits the speaker more than it benefits the listener.
52	Pliance Pliance is a response specified by the “speaker”. Pliance is always maintained by generalized social reinforcers. Pliance is passing the salt when a speaker says “pass the salt”.
53	Tracking Tracking is a listener response where there is not a direct correlation between the speaker’s utterance and listener's response. Tracking is maintained through natural consequences and generalized social reinforcement when it is a natural consequence of the response. Tracking is getting out an umbrella when someone says, “It is raining out”.
54	Milieu Language Training Model The milieu language training model is a naturalistic, conversation-based teaching procedure in which the child’s interest in the environment is used as a basis for eliciting elaborated child communication responses. It evolved from the incidental teaching model, which was developed for language delayed children.
55	Milieu Language Training Model There are four milieu teaching strategies that have been demonstrated to be effective in teaching new language skills to children with autism. Incidental teaching (teaching elaborated language in response to the child’s requests) is the primary strategy emphasized in this component of the intervention. Child-cued modeling.
56	Milieu Language Training Model Mand-modeling. Time-delay techniques. Milieu teaching in this hybrid model is embedded in the arranged environment and the conversational style taught in the first and second components of the model. Milieu teaching episodes are relatively few and carefully matched to the child’s interest and intended language functions.
57	Incidental Teaching Incidental teaching is designed to increase spontaneous speech. “Spontaneous” verbal behavior “verbal response in the absence of a verbal discriminative stimulus” Charlop, Schreibman, and Thiboeau (1985).
58	Incidental Teaching Incidental teaching uses four basic techniques. (1) student choice of activity or objects. (2) blocking access to material or event. (3) placing desired material out of reach. (4) offering students the opportunity to use existing communication skills so such skills can be elaborated or shaped.
59	Incidental Teaching When used with DTT, incidental teaching has demonstrated the ability to facilitate generalization ( McGee, Krantz, Mason, & McClannahan, 1983).
60	Incidental Teaching Incidental teaching has demonstrated superior results in acquisition and generalization when compared to DTT with reference to language training and social skills. Incidental teaching could be enhanced by a better understanding of momentary aspects of EO’s.
61	Mand-modeling Mand-modeling (Warren & Bambara, 1989) is another naturalistic training technique that involves setting up establishing operations. Letting the morning pass without food. Arranging the context, placing cookies on the table. Natural behavior occurs, child reaching for the cookie.
62	Mand-modeling Response is blocked until the child requests the cookie. If the child does not request a prompt is given. If the request is still not made another prompt is given.
63	Mand-modeling If this request is not made the teacher coexecutes the response with the child. Puts the child’s hand on the cookie. Models the response, “I want cookie”.
64	Time Delay The time delay procedure is designed to give the child an extended opportunity to respond. In a study where teachers were trained to delay 5 seconds before offering assistance or materials. A child’s failure to respond resulted in the teacher modeling the correct response for the child.
65	Time Delay Results demonstrated that time delay increased spontaneous requests for assistance. In other studies time delay procedures were found to be as effective in developing “self-initiated” verbalization in autistic children as a treatment package consisting of verbal modeling, visual cues and fading techniques, and training with multiple exemplars.
66	Teaching Complex Verbal Behaviors: Everyone must have the focus of language as the key element to the successful intervention. Language training must be incorporated into all other activities. There must be enormous amounts of trials delivered daily in the language area under a variety of stimulus and reinforcement contingencies.
67	Teaching Complex Verbal Behaviors: Staff and family must be maximally trained to carry out the treatment protocols. What approximations to accept and how to shape better approximations. What level of prompt to provide. How to fade the prompts quickly and effectively.
68	Teaching Complex Verbal Behaviors: Consistent in their implementation. Must be able to identify and prepared to optimize the opportunity to evoke verbal responses by capturing or arranging opportunities to communicate. Frequent opportunities to generalize.
69	Teaching Complex Verbal Behaviors: An understanding of novelization or generalized responding in novel environments or with novel stimuli. All types of trials must be conducted including receptive, mand, tact, and intraverbal. A curriculum matrix should be created that shows a logical progression to more complex verbal behavior. Data must be collected on daily performance and used to measure effectiveness over time.
70	Methods for Programming Identity Matching to Sample
71	Phase 1: Reliable simple discrimination. Establish discrimination of presence versus absence of the form. Scan the stimulus, Respond by touching the form, Ignore irrelevant cues, Refrain from touching when form is absent.
72	Phase 2: One trial simple discrimination learning. Within one trial with sample “A” a learner’s behavior is reinforced for selecting comparison “A”. On the very next trial a learner’s behavior is reinforced when it immediately comes under the control of sample “B”. Teach discriminations using errorless techniques. Stimuli are drawn from a large pool of different forms so learners continually acquire new discriminations.
73	Phase 3: Nonconditional identity matching, At the end of phase 3 new stimuli continue to appear on every trial within a session, the correct / incorrect functions of the choice stimuli are not conditional. Demonstration trial becomes the sample, Discrimination trial becomes becomes the comparison display.
74	Phase 4: Conditional identity matching with reversals. When “A” is sample then comparison “A” is SR+ and “B” is SR-, When “B” is sample then comparison “B” is SR+ and “A” is SR-, Initially two to three sessions are interspersed between function reversals. Gradually reduce the number of intervening trials.
75	Other Methods for Training Arbitrary Matching-to-sample Sample stimulus control shaping methods Yoked reversal (contingency class) procedure Differential sample-response procedure
76	Recombinative Generalization One of Chomsky’s (1959) greatest mistakes with criticism’s about Skinner’s (1957) verbal behavior was Chomsky’s failure to understand Skinner’s concept of the autoclitic frame. For example, Chomsky faulted Skinner for being unable to account for the word “mouses” since children wouldn’t have heard this form of the response in the verbal community.
77	Introduction Had Chomsky fully read and understood Skinner’s work he would have realized that Skinner asserts the (s) in a plural form of the tact becomes the autoclitic S frame during recombination. When the novel response is emitted it is emitted under the control of the tact “mouse” and the plural S frame thus, when one novelly tacts mouse, it may be tacted mouses.
78	Introduction While Skinner gave us the theoretical groundwork for the autoclitic frame, real research was being conducted by behavioral linguist Esper, under the title “analogy learning and miniature linguistic systems”. This research was merged by Wetherby (1978) in a chapter titled miniature linguistics systems and A functional analysis of verbal behavior.
79	How are Autoclitics Created Howard Goldstein laid out Esper’s matrix training format and has used the format to develop very rapid generalization of new words and combinations.
80	Recombinative Generalization Matrix training using object-location matrix has demonstrated with persons with retardation that after training four basic elements to establish the frame over 49 new combinations emerged.
81	Recombinative Generalization Also, given four expressive combinations of object-preposition-location, 90 untrained responses emerged.
82	Recombinative Generalization This technique offers much hope for the future of language training in retardation and autism. Other matrix patterns that have been developed are agent- action matrixes, letter word ending combinations (such as pluralizations). Matrix training has demonstrated efficacy in cross model transfer( that is transfer from receptive to expressive or cross from commenting to requesting..