FAD2 Day 4 - All About Reinforcement
Food Reinforcers Work Best For Training New Skills
Any trainer will tell you that it's best to use high value food for your dog when training new skills. It's easy to learn how to use food as a reinforcer that it is to learn how to use other types of reinforcement. It's also faster to supply a food treat and rapid rewards during training sessions when you can use food. And most generally all dogs need to eat and most love to eat. You do get the occasional dog like Azul where food has become almost aversive due to some stomach troubles he experienced as an adolescent. Now he really only works for food when we are working with another team that is using food reinforcers. Often dogs will not take food if they are stressed or over excited. Ideally we are training new skills in calm, quiet, low distraction zones where food will easily and happily work for reinforcement.
However if you are like Penny, you don't want to carry large amounts of food around with you forever so you look for other reinforcers. The trick with reinforcement is that it's up to the dog to decide what they love and what value they assign to different types of reinforcement. Check out the Yooper Paws blog on Understanding Reinforcement for ideas on determining what your dog finds reinforcing.
Most generally the first way we deliver food reinforcement is by hand feeding them the tasty morsel directly to their mouth. This is very effective when are training in low distraction environments and working on repetitions in the behavior we are training. But there are several other ways to deliver food reinforcement. Here are just a few:
In many games, you toss a piece of food away from you to get the dog to leave your side so that you can call them back and get into the position you are training again. This is often used with teaching proximity, recall, and even mat training.
Cindy likes to play a game where she holds her hand above Nick's nose or head and drops food between her fingers for Nick to catch. This works well for working on heel where your dog is watching your hand to learn the position you want them to walk in.
For dogs who love to sniff, like Azul, food can be made to be more reinforcing when Penny scatters a few pieces in the grass for him to find. This is often good to start a training session and get Azul started eating. This method also works great for training around distractions as the food gets the dog sniffing the environment around them thus taking focus away from the distraction.
There are many other ways to deliver food reinforcement. Often it helps to get creative and try out delivery methods that mimic what your dog's breed was originally designed to do. Terriers may love sticking their head inside a box or bucket to find food. Hunting breeds may enjoy searching or chasing food. Herding dogs may enjoy gathering food into 1 place and eating it all at once. Azul loves to bury his treats in the snow and dig it back out a few minutes or sometimes hours later to enjoy it.
Phasing Out ReinforcersWhile we never want to totally phase out reinforcement for our dogs, we may want to fade out food reinforcement for known behaviors and transition into social or environmental reinforcers. We also phase out reinforcement as we build up duration for certain skills such as staying on the mat. Yet there really isn't much information out about phasing out reinforcers or how to do it without slowing down your progress or seeing a behavior begin to go away because it's no longer being reinforcing for the dog.
The first thing Penny starts with is building up a delayed reinforcement. This can simply start as you use a clicker or marker word to let the dog know they just did the right behavior and it takes you longer and longer to dig the treat out of the treat bag. Over time, you slowly build up time between the marker and the treat, perhaps the dog something great like retrieves your slippers and you have to walk to the treat jar in the other room to deliver the reward. Building in a slight delay is often the first step in phasing out the reward.
Another way of phasing out food rewards is to begin substituting food reinforcement with other types of reinforcement including social reinforcers (praise, petting, etc), environmental reinforcers (sniffing, the act of moving, watching wildlife, etc), and functional reinforcers (gives the dog what they want most in the moment such as a dog that barks at a trigger often wants the trigger to move away).
We previously talked about Environmental Processing and how it's important to let dogs be dogs as well as use the environment to gather facts about what distractions and triggers might be nearby. For FAD Day 4 we are going to look at how different things in the environment can be reinforcing for our dogs if we can learn to use them in a way that encourages our dogs to continue doing the behaviors we love.
For example, one of the most common behaviors dog owners struggle with is pulling on the leash. There are multiple reasons that make learning this behavior so challenging for our dogs, but learning to use reinforcement correctly is often the biggest challenge for the human end of the leash. The dog thinks an amazing smell is coming off that tree up ahead and pulls toward the tree. The dog either drags the owner to the sniff spot or the owner puts on the breaks and dips into their training resources in an attempt to convince the dog that walking nicely is more beneficial to both human and dog. That can be a pretty hard sale, especially if all we have in our bag of reinforcers is lower value foods. This is often where dog owners slip into corrective based mode with pulling back on the leash are issuing orders to stop pulling, but as we all know these are rarely effective in stopping the pull unless you're willing to hurt the dog to get their attention which we believe is unethical and unacceptable. Another common technique owners try is to simply stop, which applies gentle leash pressure and wait for the dog to look back to them so they can move forward again together. This works for some dogs and more easily when the tree is a short distance ahead but not always when the tree is a further distance ahead so both dog and human need to repeat this behavior numerous times before reaching the reinforcement.
This is where environmental reinforcers come into play as well as functional reinforcers. The dog wants to sniff and has a basic doggie need to sniff. (And yes the human has a need to stay safe without being pulled to every distraction!) We can use that need to sniff to our advantage if plan ahead with higher value (more smelly food), produce a scent article the dog loves, or point out a possible sniff spot that is closer to the human in an attempt to get the dog to sniff elsewhere first which allows the human to move closer to the tree the dog really wants. For some dogs tossing some higher value food into the grass by the human's feet can be enough to get the dog to come back and sniff closer. I've been accused of having pocket sardines in an attempt to have smelly enough food to get Azul's interest. Yet in reality food is low value to Azul and all smells in the environment outweigh that of food. If the tree is too far away I can kick at some leaves, dirt, or sticks (or snow) with a cue, What's This?, and Azul will almost always come to investigate. Where we live, there is almost always something of interest that I use as a reinforcement for Azul for coming back toward me instead of pulling me. This is a great functional reinforcer that meets the needs of the dog by giving them what they want in a way that meets the human needs as well.
Social ReinforcersSocial reinforcers are a bit more challenging to learn to use because all dogs enjoy different aspects of being social. For a dog who loves to play and train with other dogs, Azul is heavily reinforced by playing games based concept training games when other dogs are around. Azul has always loved greeting strangers as well, but it's clear to see that he enjoys the greeting primarily for the ability to sniff the strangers shoes or legs.
When it comes to using social reinforcers it's important to make sure that all parties involved, both human and animal, approve of and feel safe, calm, & happy with the interaction. If just one party doesn't want the interaction or wants to stop the interaction, then everyone needs to back off and respect that desire. This is another huge area where dog professionals are trying to re-write the narrative to provide more human and ethical care for our dogs. No animal, human or dog, has the right to invade another animal's space without consent! And we as humans do not have the right to allow other humans to invade our dog's space without our dog's consent. Again that is a whole different lesson, but it was important to mention it here.
It is extra challenging to using social reinforcers to our advantage to increase more of the behaviors we want our dog to repeat. If we have the overly social dog like Azul, especially as a puppy, it can be a negative reinforcement if we don't allow them to greet every person and every dog they see. Yet it's also important to teach our dogs not to be rude and to respect their right to say no to the interaction. When doing Service Dog training with Azul, I had to figure out a way to communicate with Azul that during "work time" we didn't socialize and during "free time" we could socialize. This meant that all greetings followed the rule of my need to grant permission first simply to prevent Azul from rushing everyone rudely. I also had to establish certain environments where greetings would occur more often such outdoor environments, after the check out process in a store, and at the end of training sessions or group activities with other dogs. Then lastly I needed to figure out a way that prevented other people who were trying to engage with Azul from totally distracting him when he was supposed to be in "work mode" and focused on me. Basically I let Azul know that I could be way more reinforcing then the strangers who would frequently make comments about his eyes and use the words of gorgeous, handsome and cute. I did this by amping up my excitement level around these distractions, ignoring the distraction while moving away, and cuing the behavior I really wanted in that moment as part of a fun, interactive team game. This taught Azul that those often heard words were actually cues to increase his focus on me and ignore the distraction which is always more self reinforcing now because I was able to switch one form of social reinforcement (from strangers) to another form of social reinforcement (from his human) and therefore meeting Azul's need for social and team connections.
Games & Exercises
Social Events with Penny
We've already discussed puppy playdates and consent for social interactions so for my game I'm going to focus on social activities that take place in the home. Many dog owners struggle with dogs doing rude behaviors while trying to over enthusiastically greet friends and family who come into our home. This is often one of the hardest times for our dogs to feel safe, calm, and happy. For dogs that feel unsafe, generally I remove them from the area until the house guest(s) have come in and settled in allowing people to do the exciting greetings without risk of scaring the dog even more. Removing dogs to a safe space that has been conditioned with positive associations can give them a chance to sniff the new arrival from a place the dog feels safest. I will also use this management technique with overly excited dogs that need to have a moment to find their calm before greeting the guest(s). Once all parties are calmer, greetings generally go over much better.
Teaching the dog this routine for calmer greetings is not always easy! This is where we need games and exercises to help set our dogs up for success. Before we can play games, we need to first make sure the dog feels SAFE in the home and with all those who live in the home which may still be a work in progress if you recently rescued a new dog! The Fool Around Game from earlier in this workshop can come into play here with specific exercises that are set up for success. Handler 1 is the main person the dog has bonded with or their #1 person. Handler 2 is another family in the home. Start with #2 interacting with the dog playing their favorite game of tug, fetch, or whatever. Handler #1 sneaks out and remains gone for a brief moment and pretends to come home but ignores the dog and #2 does their absolute best to keep playing by amping up and getting as excited as they need to keep the dog focused on them. Repeat the games and use a barrier to prevent the dog from reaching Handler #1 if #2 can not keep their interest. When Handler #1's arrival home is no longer over exciting, switch rolls so #2 is now the person coming home. Build this game up for success with the same to people over and over again before adding in additional people that the dog already knows well. Keep playing over the course of a long period of time until you can begin to play the game with strangers (to your dog, not necessary strangers to the human.)
Once the over excitement level has changed you can also add in Mat Training here and calm, appropriate greetings.
Be A Dog/Go Sniff with Cindy
It's pretty common when your working with Service Dogs or other Working Dogs that you teach the dog when they should be in work mode and when they are free to play and participate in doggy type behaviors such as sniffing in nature. Cindy gives Nick the "Be A Dog" cue to tell him he can move away from the heel or close proximity walk and explore the environment around him.
This goes back to the dog's need to explore the environment and be allowed to interact in that environment in a way that makes them happy and/or reduces stress.
Azul likes to dig in the field and Penny has put this on cue so that he doesn't dig up the backyard but can go dig in the garden or on the hillside where the moles like to live. Penny has also taught a Go Play cue to let Azul know he can go have fun with his friends.
The Be A Dog/Go Sniff game is really easy to teach because it's based on what dogs do naturally. Start by asking your dog to focus with a few simple cues, then toss some treats on the ground and give your release works "Be a Dog!" When the dog runs after the food and starts sniffing, toss some more food down as a reinforcement. You can even spread the food out so the dog has a bigger area to explore. Since most dogs love the behavior, it doesn't take long at all for them to learn your cue and you can begin to phase out the food.