ep 90 phases of teaching something
===
[00:00:00] If you train, compete, teach, or simply love the sport of dog agility, you're in the right place. Hey there, I'm Megan Foster, creator of Fostering Excellence and Agility. Join me as we explore training, communication, handling, mindset, and all the little details that help build stronger and more confident teams, both in and out of the ring.
Whether you're just getting started or working toward big competition goals, this podcast is here to support your journey with your individual teammate. Let's get started
Hey there, this is episode 90, and today I'm going to take you through my process of teaching something. So I've tried to break this down into different phases, and so I will explain each phase and dive in a little bit deeper to each one without making this too overwhelming or nerdy. Phase one is acquisition.
You [00:01:00] have to create the behavior. You have to open up a new document in your dog's head and start filling it with the behavior that you want to see. I generally don't mind how this is done. I don't have strong opinions on if something should be shaped purely or lured or prompted or captured. I'm going for kind of the quickest route with the least amount of frustration.
Personally, I do a lot of luring and prompting and setting up my training session so that the behavior is really, really obvious rather than waiting for my dog to magically pull a behavior out of their head. Um, and then I get the behavior going pretty quickly so that the dog can offer it, but I have to set the situation up so that the behavior that I want to see is really obvious, and if it's not, [00:02:00] I do rely more on prompting and luring just to make the behavior happen.
I don't want to bake confusion, frustration, or just the dog and I staring back at each other not knowing what to do into my training. So I want to get the behavior, um, happening relatively quickly. So when the situation presents itself, however I've set it up, I want the dog to go, "Yep, I know what to do in this context, in this situation," and they do it really quickly Once they're doing that, I tend to name it, and I name it no matter if I have the finished behavior, because maybe the behavior's not that complex, or I have a piece of the behavior.
So a finished behavior might be wrapping a cone or wrapping a wing one time. I'm, [00:03:00] that... I might be able to acquire that behavior in one or two sessions, and it never has to change. So that's the finished behavior. It has to be generalized, of course. But that is the whole thing. More behaviors won't be added to it.
Versus weave poles, I'm also really comfortable naming, um, just two poles if I'm using a two by two method. I'm really comfortable naming a wide open channel the cue that I will expect to use when there's 12 poles or 60 poles or completely closed poles. So I'm very comfortable naming things sooner rather than later in the process, rather than having a general do the behavior cue that I then have to change later on.
What matters most to me in the acquisition phase is the amount of latency that is there. So when I make the behavior available to the dog, either [00:04:00] via the training situation, the setup within the training space, or the cue, I want them to start that behavior very quickly. I'm not going to move on to the next phase until I have very low latency.
I can tell that the dog hears that cue or sees that situation and says, "I got this. I know what to do." So I'm also very particular about how much arousal I layer in at this stage, because it primarily depends on how well the dog self-regulates outside of training and inside of training. So one thing that I wish I had done differently with Sprint, and I will do differently in the future, is to not start a training process that the dog cannot emotionally handle.
So even if the dog is physically capable of doing the skill or even [00:05:00] mentally capable of doing the skill, but they aren't able to emotionally regulate throughout learning that skill, I'm gonna put it off until they can. So this is going to be very dog dependent and also skill dependent
Once I have something on cue and the dog views that cue as a green light to immediately start performing the behavior, I move into what I call basic generalization because I want to spread the effects of that learning that they showed me about acquiring this new behavior on cue. I want to spread that experience to a variety of contexts and situations.
So this is going to include handler situations because in agility my physical cues are both very helpful to them, but [00:06:00] also a massive distraction to a lot of the behaviors that we teach and need them to do independently of how fast we're running, which direction we're running, how close we are to them, how far we are to them.
So in basic generalization, I'm making sure that they can still perform this newly acquired behavior regardless of where I'm positioned and what my motion is doing within reason for the specific behavior or piece of behavior that I have now put on cue. I'm making subtle picture changes to the environment especially.
I want to kind of take my dog out of their typical training space, which for a lot of my basic beginnings of training something, it happens in my living room. So being able to do it in my living room isn't going to help me much for a competitive agility [00:07:00] career, so I need to get the dog doing this behavior in a bigger variety of environments.
Um, so it might be a different room in my house, it might be front yard versus backyard, it might be the local park, it might be an empty parking lot. It might be adjacent to a dog park or a kid's soccer game. Either way, I want to make sure that this behavior can happen in a variety of environments that my dog is already fairly neutral and comfortable with.
Environmental changes include other people, dogs, noises. I've talked in the past about when I'm adding distractions into training, I want my dog to already know how to respond to that distraction before I'm adding it to a specific training task. So do keep that in mind, that if the [00:08:00] dog doesn't know how to react when a new person enters the training space when they're doing nothing, they're going to have a very hard time knowing what to do when they're doing this newly acquired skill.
During basic generalization, I'm also changing things up within my reinforcement strategies and procedures. So if I primarily used a pre-placed reward for acquiring the behavior, I'm going to make sure that the behavior still exists if that reward is no longer pre-placed, if the reward can be in my hand.
Same as if the reward was coming mostly from my hand, and when I'm training something, I have food in my hands all the time, I'm going to make sure that that behavior can still be performed if the food is in my pocket or if the toy is not visible. So we're gonna make small changes to the reinforcement strategy just to make sure that it wasn't accidentally [00:09:00] attached to the behavior that your dog learned.
This is where we can also play around a little bit with arousal mobility. Different reinforcers and different reinforcement strategies are going to impact the behavior differently. So if you were using a low-value, low-arousal-level reinforcer to acquire the behavior, increasing the arousal via the reinforcement strategy is going to help you to generalize the behavior to more high-arousal situations later.
What's most important in the basic generalization phase is that your cues are clear and salient because we need to make sure that the dog is attaching the behavior to the correct thing. If the dog can only do the behavior if, you know, you're in a certain place or they're in a certain room or [00:10:00] there, there's a certain reinforcer used, generalization is not happening.
So that will mean that the dog has attached the behavior to those specific things rather than the cue that you are giving them. And I don't want to move on until I know that the dog has learned the cue that I want them to learn in order to start performing that behavior
Before I move on into more advanced generalization and increased levels of arousal, I need to check on the level of stimulus control I actually have over this behavior. Stimulus control just means that the behavior is only occurring in the presence of the cue that you attach to that behavior and not when that cue isn't given.
So what I've learned a lot about [00:11:00] training, and especially with the increased importance on verbal cues in dog agility, is that not doing the thing is just as important as doing the thing. Let me explain. If you approach a jump but you haven't asked the dog to do it And they just leave you and offer that behavior.
That is a lack of stimulus control over jumping. And this is similar to when you pull out a bag of food and your dog just starts throwing behaviors at you all over the place. That means those behaviors are not under the stimulus control that you believe them to be. In fact, you pulling out food is one of the cues that they understand to start throwing behaviors at you.
So it really c- does come down to is the behavior on the cue you think it is, or are there additional cues [00:12:00] that can make this behavior happen? And I don't wanna scare anyone with the idea that you need 100% stimulus control over all of your behaviors all of the time before you can move on. I don't think that that's necessarily true, but I do think that in more advanced training, when we come to a sticking point, there's too much failure, there's too much confusion, there's too much frustration, I truly believe that it does come back to stimulus control, and maybe we need to go back into more basic training, fill in some of those gaps surrounding stimulus control before continuing on with harder and harder things t- that we're asking of the dog.
Uh, so again, understanding the whole cue picture is so important for understanding stimulus control. So not only does your physical cue [00:13:00] need to be clear and consistent, the verbal cue needs to match that physical cue if there's a verbal cue for it, and the obstacle you are asking them to engage with when they hear and see those cues needs to be available to them in order for them to perform the cue The dog needs to perform the behavior immediately when they perceive the cue, which again means the obstacle needs to be available to them And a form of stimulus control is that the dog does not offer any other behavior when you give this specific cue.
So if you ask the dog to go to the backside, but they go to the front side, that's offering a different behavior for your backside cue. And so if that's happening chronically, one of the [00:14:00] things that we should look at is do we actually have stimulus control over both of those cues, or is there more basic training we can do to improve the stimulus control?
So once we feel pretty comfortable with the amount of stimulus control that the dog has, and really I just want to know that the dog can wait for the entire cue picture before starting the behavior, rather than starting the behavior before they've been given the entire cue picture But then we can move into advanced generalization and discrimination.
So discrimination is the tendency for a behavior to happen in one situation but not another. So what... The dog should take the tunnel when the [00:15:00] situation calls for it, and the cue picture gives them the green light to take the tunnel versus they should take the contact obstacle when the situation presents itself, and the cue picture gives them the green light to take the contact obstacle.
Okay? That is discrimination. Stimulus control is part of discrimination training because not doing the thing when it is not being cued versus doing the thing when it is being cued is kind of level one of discrimination training. But I can also continue to make more challenging distractions a part of my advanced generalization.
I can start increasing the arousal level surrounding the situations and questions that I ask, uh, for these behaviors in, and I can continue adding layers [00:16:00] of understanding now that I have enough reinforcement history of that behavior happening when I cue it and not happening when I don't cue it. In my verbal cues program, I really break down this phase, this advanced generalization phase, into, like, five parts about very slowly layering in the different discrimination tasks to avoid, uh, too much guessing, frustration, and confusion And then, and maybe even before, right?
These are different phases, and I can't say for certain that I would just go straight through all the phases before I would say, "Okay, we're ready to put this behavior into competition or sequencing or coursework." It's not [00:17:00] really that, because I don't believe that every behavior has to go through advanced levels of generalization if your team doesn't need it to.
So you can move back and forth through the phases, and you can skip over an advanced phase until you need it, things like that. But beyond advanced generalization is testing and polishing. Chaining things together is the best form of testing how a behavior is holding up under pressure. When we are chaining behaviors, we are sequencing a series of related behaviors, the last of which is being reinforced.
So this means that the second behavior you're asking for has to have a reinforcement history behind it big enough to [00:18:00] reward the first thing that you asked for. And the third thing has to have a reinforcement history big enough to pay for the second behavior, and so on. So generally, when we start chaining behaviors together, we will uncover what needs more work, because we'll start to see small mistakes or a lack of motivation or a, just a misunderstanding, and that's generally directly tied to the reinforcement moving and the behaviors relying on the reinforcement history of other behaviors to keep things tidy and neat and clean.
So remember that each behavior we're chaining together has a cost to the dog as well, either mentally, physically, or both. So when we're chaining [00:19:00] multiple expensive behavior, behaviors together, we might see more mistakes pile up than if we're chaining less expensive behaviors together. So those are things to think about when you move into a testing phase
So to quickly recap, the main phases of teaching something is to, one, acquire the behavior and get it named, get it on a cue. Two, do some basic generalization. Make sure that your dog can perform this behavior in a variety of contexts and situations. Three, make sure you have a reasonable amount of stimulus control over this behavior before moving into, four, advanced generalization and discrimination.
And [00:20:00] any time you feel like you would like to, you can test a behavior and continue polishing that behavior. So it might even be that I get a dog through basic generalization, and I want to test that behavior with a novice level sequence of behaviors and gather the information that I need before moving into higher levels of advanced training, and then test it again later on.
Okay? So that is my process, the framework that I try to take every behavior through, and I would love to know what you think. So come on over to the Fans of FX Agility community, and let's talk about it
Thanks so much for listening. If you enjoyed this episode, I'd love for you to leave a review, share it with an agility friend, or [00:21:00] connect with me on social media. Your support helps more handlers and teams find the podcast and be part of this community. You can also find all of my training programs, coaching, and resources at fxagility.com.
Remember, excellent agility is built one training session at a time, and every small step forward matters. Until next time, happy training
