Showing is often faster than telling. Modeling means demonstrating a behavior so the learner can imitate it, an efficient way to teach a new action, a social skill, or a staff procedure. It only works, though, when the learner already imitates and when the model is set up so the learner actually notices it.
Why it matters
A clear demonstration can save many shaping trials. Modeling serves as a response prompt in discrete trials, a stand-alone procedure for social and daily living skills, and one of the four parts of behavioral skills training for RBTs and caregivers.
Practice items for this task check three things: whether the learner has the imitation repertoire modeling depends on, which model arrangement is most likely to be imitated, and which type of modeling a vignette describes (live, video, self, peer, or point-of-view).
Key ideas
- Modeling is the procedure; imitation is the learner's response. You present a demonstration (the model) and the learner produces a response with the same form, shortly after, because of that demonstration.
- Imitation is the prerequisite. If a learner does not yet imitate, extra demonstrations will not help. Teach imitation directly first: start with a small set of simple models, prompt and reinforce matching responses, and probe untrained models to check for generalized imitation. If the learner still fails most untrained probes, add new, more varied imitation targets (different actions, materials, and people) before relying on modeling to teach new skills.
- Models can be live or symbolic. A live (in vivo) model is a person demonstrating in the moment. A symbolic model is presented through video, pictures, or other media. Video lets you repeat the exact same demonstration and edit out errors.
- Know the main video formats.
- Video modeling: the learner watches someone else, often a peer or adult, perform the skill.
- Video self-modeling: the learner watches edited footage of themselves performing the skill successfully, with prompts and errors removed.
- Point-of-view video modeling: the video is filmed from the learner's visual perspective, often with only the performer's hands in view.
- Model characteristics matter. Learners are more likely to imitate models who are similar to them (age, setting, background), who are competent at the skill, and whose behavior the learner sees reinforced.
- Arrange for attention and practice. Make the model salient, keep the demonstration simple or break it into steps, give the learner an immediate chance to respond, and reinforce correct imitation.
- Fade the model. A model used as a prompt is a temporary support. Move from full models to partial models, or add a delay, until the natural cue controls the behavior.
- Program generalization. Use more than one model, setting, and set of materials so the skill is not tied to one demonstrator.
Definitions
- Modeling: presenting a demonstration of the target behavior so the learner can imitate it.
- Model: the demonstration itself, an antecedent stimulus with the same form as the response you want.
- Imitation: a response that matches the form of a model, occurs soon after it, and is controlled by it.
- Generalized imitation: imitating new models that were never directly trained or reinforced.
- Video modeling: a symbolic model in which the learner watches a recorded demonstration of the skill.
- Video self-modeling: video modeling in which the learner watches edited footage of their own successful performance.
- Point-of-view video modeling: video modeling filmed from the learner's visual perspective.
- Behavioral skills training (BST): a training package of instructions, modeling, rehearsal, and feedback, repeated until the trainee meets a mastery criterion.
Examples from a session
Modeling across ages and settings
- Clinic, model prompt: During play, an RBT says, "Do this," and taps a drum. Malik, 4, taps it right away and gets a few seconds with a favorite spinning toy. Over later sessions, the RBT shows only part of the tap (a partial model) and then fades the model entirely.
- School, peer video model: Nina, 12, watches a short clip of classmates greeting each other and asking to join a lunch table, then practices with a peer before lunch.
- Community, point-of-view video: Jorge, an adult learning to ride the bus, watches a video filmed from a rider's point of view: hands tap a card, the camera moves to an open seat, and a hand pulls the cord at the stop, just as Jorge will see it.
- Home, BST with a caregiver: A BCBA describes how to deliver a token right after a request, demonstrates it twice with the client's brother, has the parent practice, and gives specific feedback until the parent delivers tokens correctly on several trials in a row.
- Non-example: An RBT says, "Tap the drum," and Malik taps it. No demonstration came first, so this is instruction following (task G.10), not imitation of a model.
Common mistakes / exam traps
- Repeating models with a learner who does not imitate. If no imitation follows several clear demonstrations, the fix is imitation training, not more or louder demonstrations.
- Assuming video solves an imitation deficit. Video models still require imitation. Video helps with consistency and attention, not with a missing prerequisite.
- Leaving the model in place. A model prompt that is never faded can leave the learner prompt dependent.
- Calling BST complete without modeling. Instructions plus rehearsal plus feedback is missing the demonstration step.
Exam trap
Imitation deficit vs. model problem: when a vignette says modeling "isn't working," check the prerequisite first. A learner who does not imitate even simple live models needs imitation training. A learner who imitates live models but not this one needs a better model: check attention, similarity, and whether imitation is reinforced.
Exam trap
Video self-modeling vs. point-of-view video modeling: self-modeling is about who performs (the learner is the model). Point-of-view is about camera angle (the scene appears as the learner would see it, often showing only hands). An edited clip of the learner succeeding, filmed from across the room, is self-modeling, not point-of-view.
Quick check
Ana, a BCBA, has been teaching Theo, 6, to wave hello by waving and saying, “Do this.” Theo now waves on every trial. Ana starts lifting her hand only partway before Theo waves. What is Ana doing?