Imitation Lecture Notes

Imitation | Lecture Notes ● Imitation Training ○ Imitation refers to a correspondence relation between an SD and a behavior that share formal similarity and occur close in time to one another. ○ Many things we learn growing up were modeled for us and we imitated that model. ● Components of Imitation ○ Presence of a lot of stimulus ■ Which is presented in an effort to evoke the imitative behavior ○ Immediacy ■ ■ The imitative behavior has to immediately follow the model stimulus Immediacy is very important and as a defining feature of imitative behavior. ■ In technical terms, the presentation of the model and the occurrence of imitation should be in close temporal relation to one another. ■ In layman's terms, this just means that the imitative behavior has to immediately follow the model. ○ Formal Similarity ■ ■ The model and the behavior must have formal similarity Meaning they need to physically resemble each other and be in the same sense mode, such as physical behavior. ■ ● I.e., a physical model produces a physical response. ● I.e., a coval model produces a vocal response Formal similarity refers to the model and the behavior having a physical resemblance to each other and being in the same sense mode. ■ In layman's terms, this just means that the model and the imitative behavior have to be identical to one another. ○ Controlling Variable ■ The model must serve as the controlling variable for the imitative behavior ■ ■ The model has to be the SD NOTE: If youre using a prompt along the way or some vocal instruction, then this is no longer an imitation response. ■ After the model evokes an imitative behavior, that behavior comes into contact with contingencies of reinforcement. Tthat's when these contingencies become the controlling variable for the discriminated operant ● Imitation Models ○ The models that we use in imitation are antecedent stimuli that are meant to evoke the imitative behavior ○ The model needs to serve as your controlling stimulus for that imitative behavior. ○ These models can be planned or unplanned ■ Planned models are prearranged antecedents that help learners to acquire new skills or to refine the topography of existing skills ● ■ I.e., yoga class - yoga teacher instructs you how to do a pose Unplanned models are antecedent stimuli with the capacity to evoke imitation but they werent specifically programmed to evoke that imitation ● ○ I.e., see your sibling playing with a toy and you copy him. A novel model evokes a similar behavior in the absence of a history of reinforcement for that behavior. And so this is the unique defining feature of imitation. ○ One of the important features of imitation is that in a typically developed organism, imitation is present without having to specifically have been taught. ● How to Conduct Imitation Training ○ The objective of imitation training is to establish a repertoire of generalized imitation. ○ ○ This basically just means teaching the learner to do what the model does. Before you start training, you want to assess and teach prerequisite skills required for imitation to occur. ■ This includes joint attention, which means that your learner can make eye contact with you and track your movements, and it includes compliance. ■ Yur learners should follow your instructions when you tell them to look here, copy me, do this. ○ You want to minimize problem behaviors that may interfere with training. ■ It might be necessary for you to address the problem behaviors before doing any type of formal instructional training like skill acquisition training. ○ Next, you want to select your models, so pick about 25 gross and fine motor examples and use one model at a time. ■ ■ Don't do multi-imitation just yet. Present these models one at a time. In this first phase, in the pre-training phase, we're just going to be evaluating to see which of these models the learner can do and which ones are problematic for the learner. ○ Next, we're going to go to our pre-testing phase, where we will determine if the individual can already imitate some of our models. ○ ○ The pretesting procedures are to have the learner be in the ready position. If an object is going to be used, if you're doing object imitation, you should have that object in your hand. ■ For example, you might tell the learner to tap the pen on the table, or you might say copy me, and do this, and expect the learner to copy that. ■ If they have a correct response, you want to praise that. But you also are just praising for cooperation, and you don't want to associate reinforcement with these responses at this point because we're just probing for what skills currently exist. ○ Next, you want to sequence your training model, so you want to arrange these from easiest to most difficult. ○ The reason for this is because you want to set your learner up for success during your training procedure. ○ Initially, you want to select models for training that were imitated correctly on some, but not all, of your pretest trials because then at least you know that this might be in their repertoire, but it may not be fully formed, or the typography might look a little bit different than what I currently want it to be. ○ Next, you want to teach the ones that the learner has approximated but done incorrectly. And finally, you want to teach the ones that were incorrect, or difficult, or the learner couldn't do at all. So this brings us to actual imitation training. Your first step is that before each imitation training session, you're going to redo a reassessment. ○ You're going to take about five to 10 of these and evaluate the learner's current performance level to determine whether or not they know it at the start of your session. ■ So that might look like this, do this, do this, do this, and you just go through all of those motions. ○ The first three that are incorrect are the ones that you are going to pick for training. ■ You're going to present these three times in a random order. ● If the learner performs them correctly, then you're going to take them out of your training sequence. ● If they're done incorrectly, you keep them in your training sequence, and you pull the next one from your pre-training list into your training sequence. ○ For training, you want to use repeated presentations of one of the three models from this pre-assessment period. ■ The one that you should start with is the one that had the highest rate of correct responding or the one that was easiest for the learner to kind of pick up, and maybe they got it right some of the time. ● Continue until the learner responds correctly five consecutive times, and you can use your prompts such as physical guidance. ○ Physical guidance will probably be the one that you use most frequently in imitation training to guide the response. ○ ○ You want to gradually fade your prompts using your graduated hierarchy. In the post-assessment, you are evaluating how well your learner performs the skills that you just taught them ● Pre Assessment & Post Assessment ○ You have a pre-assessment right before you train and a post-assessment right after you train. ■ In this session, you're presenting all of the models within that training session, and you're evaluating them to see if they're getting them correct or not when they're being probed at the end of the session. ● Assess for Generalization ○ You want to probe for imitative behavior and use this to assess for generalized imitation. ○ While we're doing imitation training, we often have a target list of the behaviors that we're teaching for the person to imitate. ○ Generalized imitation is whatever behavior I do; you should be able to imitate that. ■ So we want to probe for this generalized imitation by selecting five nontrained novel models and checking for the occurrence of imitation with those models. ■ We do this at the end of each training session or intermix them into our training sessions. ■ You don't have to use pre-assessment procedures in this situation. ● Guidelines For Imitation Training ○ ○ Keep your sessions active and short. ■ You don't want to have this be long and drawn out for like 20, 30 minutes. ■ You can keep sessions to like maybe five minutes, at most 10 minutes. Reinforced, prompted, and imitative responses initially, but move to only reinforcing correct independent responses over time. ○ You want to make sure to pair social reinforcers like praise and attention with tangible reinforcers or edible reinforcers. ○ ○ If progress breaks down, you might have to back up a step. You always want to keep records of your progress, track a learner's performance, fade out your verbal prompts and physical guidance as soon as possible. ○ Remember that the controlling variable in imitation is the model, not any type of vocal response or physical guidance. ○ If progress is breaking down and you're noticing that a target is very difficult for a learner to use, then you might want to consider using a shaping program in addition to the imitation.