Here’s something interesting I learned while working on a talk I’ll be giving at the Annual Conference of the Wisconsin Association of Behavior Analysis on August 15th in Madison. My talk is “Creating Harmony Between Dogs and Special Needs Children,” and it involves discussing the benefits to the family of having a dog, but also the risks to the dogs that need to be addressed and minimized as much as possible. As we all know, even parents of typical children sometimes struggle with interactions between their children and their dog, and things can be even harder for parents of children with special needs.
While working on my talk I read a research paper that is relevant to dog training in general, even though the case study was about changing the behavior of a child with Autism Spectrum Disorder, or ASD. A 6-year old boy was “inappropriately touching” the household dogs (grabbing and pinching the dog’s anus, oh dear). Needless to say, this was both unfair to the dogs and dangerous for the child.
The therapist attempted three different types of intervention: 1) a verbal reprimand if the behavior occurred and attempts to physically block the behavior from happening, 2) DRA, or “Differential Reinforcement of Alternative Behavior,” in which the child got a jelly bean for touching the dog appropriately and 3) DRO, or “Diff’l Reinforcement of Other Behavior, in which the therapist gave the child a chance to choose the reinforcer before each session, and then told the child “If you don’t touch the dog’s bottom, when the timer goes off (10 sec), you’ll get X (chosen reinforcer) afterward”. Appropriate touching was verbally reinforced as it was in DRA, and inappropriate touching received a verbal reprimand “No touching. That means no X. Let’s try again.”
No surprise to dog trainers, verbal reprimands had little effect. The DRA method (reinforcing appropriate behavior) resulted in an increase in appropriate touching BUT no decrease in inappropriate touching. However, DRO (don’t grab dog’s butt for 10 seconds and get a reinforcement of your choice) eliminated inappropriate touching completely. You can read more details in Bergstrom et al, 2011, Research in Autism Spectrum Disorders 5, pg 218-221.
Note that there were at least 2 variables that differed between DRA and DRO: In the latter the child got to choose the reinforcement (rather than being given one, a jelly bean, that the therapist knew he liked). Second, in DRO the reinforcement came after 10 sec of not doing something, versus getting a reinforcement for performing an alternative action. This makes it a bit difficult to sort out which factor had more influence on the result (remember, this was a case, not an experiment). My guess is that both factors were significant. Certainly it is critical to understand that, just like us, dogs want different things at different times, whether it’s being let out to pee and sniff the sidewalk, get a piece of bacon or get to play ball. One of the key distinctions of good trainers is the understanding that we need to get into our dog’s heads as best we can and guess what it is that the dog wants at any given time. That said, I suspect that reinforcement for the absence of a behavior might also have been significant.
If you’re a learning theory expert then you are well versed in the term Differential Reinforcement, (DR) including DRI (where I = Incompatible behavior) as well as DRA and DRO. Most dog trainers use these concepts all the time, just without the formal labels to them. I bring it up here because so much of family dog training includes stopping a dog from doing something you don’t want it to do as much as teaching it what you do want (which should go hand in hand, right?) I teach incompatible behavior all the time, (sit when greeting people at the door rather than jump up), but we all know that some behaviors in some dogs can be challenging to turn around. I haven’t used DRO as described above very often with my own dogs or with my clients, however I did have good luck with it when Willie was obsessed with herding Sushi. Reinforcing incompatible behavior didn’t help, if anything, it made it worse because everything became a secondary reinforcer. Once I figured that out I could manage it reasonably well, but before Sushi left I had the best success with reinforcing a lack of herding for X amount of time (inspired by a seminar with Ken Ramirez, by the way). It’s not an intuitive method at all, and I suspect it’s best used only in specific circumstances. I’d love to hear your own thoughts about your own experience with it, or with dealing in general with behavior that is difficult to extinguish.
MEANWHILE, back on the farm: Even more rain! Whoopee, almost an inch in the last 2 days. It’s probably too late to save my pasture, it appears to be dead and gone; but maybe not, said Trisha the optimist? Maybe, just maybe, it’ll come back? I’ll give it another week or so before deciding to plow it up. That would be an ouch. I only have one other small pasture and if I replant I’ll have to do it in fall and then keep the sheep off for a long time afterward. Would have to buy lots more hay, and that’s not so easy to find right now.
Speaking of sheep, I realized it’s been awhile since they had the spotlight. About time I’m sure they’d say. Here are two of the ewes I call the “camel sisters:”
They are just visiting, to help me and Willie work flighty sheep.
Here’s a question for you: What’s wrong with the picture below?
Did you guess that sheep shouldn’t be eating hay in Wisconsin in the middle of summer? Yup, you’re right. The flock is eating their winter’s food and has been since late May. Maybe I’ll have to put them on a diet like I am?
Laura says
Thank you for this wonderful post. It’s a reaffirmation of how effective positive reinforcement training can be and during my dog’s puppy days, I often had to defend those methods from friends as they were very different from how dogs are conventionally trained.
My dog has taught me many lessons, but one of those is that if he knows what I am talking about, he tries very hard to obey. I think we’ve come to trust one another in that if he does something else, I try and give him the benefit of the doubt that I didn’t explain myself well and I truly feels that he does his very best to make out my confusing speech/signals. One example of this is getting on the bed. I wanted him to only get up when invited and to jump off when asked. When training these things, it can take a little while before dogs understand what exactly you’re trying to say is acceptable. He’d get on the bed, but he wouldn’t get off. He knows the word “off” in the context of keeping 4 paws on the floor, but I realized he probably wasn’t able to transfer that to jumping off the bed. To try and clarify what I wanted, I made it into a game. I’d ask him to get up and praise him. Then I’d say “Off!” in a really cheerful voice and praise him when he ended up jumping off. Repeat. He really seems to understand both now and I haven’t had problems with him jumping on the bed when I’m in it and he’s not invited, but when he is there, even in his most comfortable state, when I say, “off!” cheerfully, he seems to think the game has commenced and jumps off with his tail wagging. Glorious!
Beth with the Corgis says
Jack was raised as an only dog. He was crate-trained, but when we brought Maddie here he found it over-stimulating to hear me training Maddie in another room (“Sit? Sit! I KNOW that! I can do it, I can do it!!”). And so he would bark.
I successfully trained him not to by using the method you describe; I would periodically run back in and give him treats if he was quiet for a few seconds. It worked so well that I can now play with Maddie with the chase-it (NOT a two-dog toy!!) in another room while he is crated and he will rarely whine and almost never bark. It helps if he gets his turn first so he’s tired.
I want to throw out there, though, that if you are careful when a dog is a puppy, you can use “no” as simply the opposite of “yes” and not as a scold. I think it helps dogs to have a word that means “that is not at all what I’m looking for” so they redirect their efforts. I know that I like to be told when I’m on totally the wrong track. The nice thing is, you can use it then to interrupt a behavior if it’s incessant, to give you the opportunity to reward. I did this initially with the barking, though I could quickly move on to waiting for him to offer short quiet periods after that. So the initial session would involve me gently calling “Jack, no” from the other room and almost instantly saying “Good boy!” when he stopped and running in and treating. If I had to wait at first for him to be quiet on his own, I might never have had the chance to reward, since barking is self-fulfilling, especially in some breeds.
Jett says
My older Aussie, Kiva, was a lunger/barker when he was aroused. I used to sit in flyball class and reward him for doing anything but lunge/bark when dogs were running. Which, to my coaches and classmates, looked like I was just handing out treats for “nothing” ;-). If he went over threshold, I would take him out of the room for a few minutes but mostly I just rewarded him for periods of self-control. He still barks at flyball, but now it’s excitement for the game, not over-arousal.
I am doing similar training with my younger Aussie but sometimes I forget I “should” be rewarding “nothing” and instead ask her for other behaviors to reward (DRI, if I understand it correctly). Hmmmm. Her behavior is improving but perhaps I should put a bit more emphasis on DRO with her.
Thanks for the food for thought!
Ron Watson says
Excellent piece!
The DRO you outlined here seems to dovetail nicely with “Duration as Function of Math” – teaching duration by structuring your rate of reinforcement at a level guaranteed to keep the dog on task.
I’d probably add a Dismissal cue to this and shape, capture and reinforce the release and reorientation and then funnel that into the DRO, not unlike I would a duration behavior.
Super interesting that the DRO and DRI seem to have the same end result, but entirely different approaches. Again the intent of the handler has great bearing on performance.
Thanks for this, great food for thought…
Peace~
FidoPhoto says
Fascinating case. Thanks for sharing it with us.
I have a 5 month old Golden puppy that really really enjoys chasing and mouthing my 5 cats. At first I ignored this and, duh, the behavior increased rather than decreased. Then I tried scolding, even though I was pretty sure it wouldn’t work (and it didn’t) and then I went to Ken Ramirez’s seminar last weekend and am now doing DRO/DRA/DRI (I’m really not sure which one I’m doing).
Basically, I click and treat every time he looks at one of the cats and doesn’t move. We’ve been doing it for 5 days now and I’m not getting much done because I’m watching him all the time, but the chasing and mouthing is WAY down. I try not to let it happen at all, but he’s gotten 3 short episodes in in 5 days.
Not surprisingly, he goes for the cats if I’m not in the room with him. The goal is for him never to do it so I guess the next step is to set it up so I can see him but he can’t see me and increase the time he looks at the cats and doesn’t move before I click (right?). I’ll go back and review what you did with Willie and Sushi and see what more I can do. I also try to tire him out with exercise and training before giving him access to the cats. Oh, and I tried playing with him when he wants to play with the cats, but quit doing that because it just wound him up and then he wanted to play with the cats too so we went back to food.
Beth with the Corgis says
I just had a thought. In the example of the child given above, isn’t the third scenario really a combination of positive reinforcement and negative punishment? Because the child has chosen the reward, and therefore has visualized the reward and sort of already had sampled the feelings that come along with having the reward…so not getting the reward is akin to taking away something desired by the child?
When my dogs are in a situation where they normally get a treat, I’m holding the treats, but I don’t give them out because of bad behavior, I consider that a sort of punishment (and of course if they then earn the reward, it’s reinforcement). Wouldn’t that apply to the scenario above as well?
Cacky Vincent says
I have found that training is so much more effective if we teach the dog the concept of “this works (positive reinforcement) and that doesn’t work (verbal reprimand)”. Just another way of looking at DRO.
Kat says
I think I need a book reference to help me get the DRA, DRO, DRI sorted in my head. I’d love to be able to figure out how to reward my damaged dog for not reacting to the neighborhood stimuli that bother her. I’ve gotten her to the point where inanimate noises don’t get more than ear flickers but raised voices, unpredictable neighbor, and barking dogs are all but guaranteed to send her over threshold. We are to the point where they don’t immediately send her over threshold and I reward her with continued ball play for as long as she will stay and play ball. Unfortunately, there comes a point sometimes very quickly but occasionally longer when it’s too much and she can’t focus on anything except trying to make the noise/behavior stop. I’ve tried releasing/removing her the instant I hear something that will set her off but in her mind that seems to be encouragement for her to go bark fiercely and behave aggressively.
Really, if I could get her to the point where she could ignore the raised voices (sadly we have neighbors who seem unable to converse at any volume below shouting) and the barking dogs (an ill mannered pack of chihuahuas across the street and a very aggressive Golden next to them) she could have a pretty good life here.
Abbie says
This post is so great and definitely a topic on my mind. When I decided to start the shift from teaching preschool special ed. to dog training, I didn’t think very many of my skills would be transferable, but it turns out much of it is in fact transferable. I run a positive reinforcement-based classroom and in working with kids with Autism was already quite familiar with Applied Behavior Analysis and some of the other key parts to creating a behavior change program. I would love to hear your talk, too bad it’s in Wisconsin!
Lisa W says
I think I actually understand what you’re writing about. Not the alphabet soup of terms, but the concept (I hope). We just had our appointment with a great vet behaviorist for our anxious and fearful dog, who was doing fairly well with a lot of desensitization and conditioning and activities and some socializing, until she got injured and has been in confinement and rehab now for four months or so (with several more to go).
As her world got smaller, her fears got larger. Not good for her or for healing. Anyway, as the vet behaviorist was explaining to us some new techniques called relaxation blocks and using “look” a lot more, and treating for the smallest signs of reaction to something before it became a reaction, and treating her for just having a nice quiet leashed walk in the yard, and praising for being relaxed and calm. Looking for the calm, almost-at-ease behavior and acknowledging/praising that (and for right now ignoring other behaviors). As we were listening, it hit me like a good epiphany does that I had been spending a lot of time and energy getting Olive to do things (safe tricks, PT exercises, not barking, etc) so she wouldn’t feel so pent up, and what we also should have been doing was really acknowledging her good efforts to be calm, relaxed, and on occasion, happy. It isn’t intuitive because it’s hard to realize the absence of undesired behavior needs reinforcement. If I’m not doing something, it’s not our normal response to praise/reward that. We tend to praise/reward things we asked for that are done well or things we ask not to be done. Wow. It rocked my world. Now if we can implement it successfully.
Jeanine Renzoni says
Thought provoking topic, although I think you’ve all been using DRO a lot more than you’ve realized…rewarding wait/stays or quiet or four on the floor , or heeling in reinforcement zone anything with maintenance/duration involved because the reward can be for continuing to do nothing or non-doing, when most dogs would like to be doing something new or doing something that has gotten attention in the past. I think its interesting that the verbal reprimand remained in the 3rd scenario despite it being totally ineffective as per the 1st scenario’s results. Often, it seems that the verbal reprimand is tied into the excitement of the non-preferred behavior.
On the management/prevention side, I would think that there would have been a bottom protector added to the dog so that pain, ick factor and reactions could be lessened or even eliminated.
The DRA portion of the case seemed not exciting/interesting enough…appropriate touch, verbal praise and a jelly bean. Whereas there would have been much more reaction to the undesirable behavior by all parties. I think that when you’re trying to offer an alternative behavior the generated interest in the alternative needs to be more on par with the level of excitement and duration of the undesirable behavior. And maybe the alternative needed to be more specific and detailed, because certainly the undesirable behavior was quite focused.
The set timer part of the scenario has potential flaws because if 10 seconds is a long time based on prior frequency (I would surely hope not!) then the reward for not doing it may never be met, and if it is too short, well then the behavior might not improve enough which 10 seconds is not enough plus given some wildness/distraction would the behavior resume? I would think they would adjust the time based on previous performance and make it vary to ensure likely success. But a clock is good to have to reference time passed. But the ding! A sound marker, definitely adds to the excitement of this approach.
Marcia in NorCal says
Re: comment from Beth with Corgis, about using a “no” as the opposite of “yes” to providee feedback that “this is not the behavior I want” … YOU BET! My experience with my own dog and a others has been that it is absolutely appropriate and helpful to signal that X is not what will get you a reward. I use “nope” (it is really, really hard to put a lot of anger into “nope” whereas we all too easily add unnecessary energy to “no”) and a friend of mine uses “too bad.” Like anything else, it takes time for these alien sounds to gain meaning for the dog, but the meaning does develop.
jackied says
Is there not the risk that the dog will back chain though? That is what has happened when I have tried rewarding each of my dogs for a period of not barking. They deliberately bark and then stop. If the treat doesn’t come quickly enough for them, they do it again… not what I wanted at all.
How would using DRO for reactive behaviours differ from counter-conditioning? Bigger gaps between the treats? Or is that an inapproprite use of DRO anyway?
chloe says
@ Jackied timing is everything when marking a behavior. What you describe is very common and is a sign that what you clicked/marked/yessed…was not the silence but the combination of bark + silence.
em says
Without knowing what it is called, I used something like DRO to teach Otis not to bother us at the dinner table, which he really, really wanted to do (and who can blame him, when all that delicious human food was right at eye level?). I began by rewarding him for lying on his bed across the room, then I’d ask him to wait and periodically pop up during dinner to reward him for staying there. At first, we had problems with him back chaining- popping up just to get us to reward him for lying back down, so we stopped rewarding “breaks”. We’d reward him for the initial down, ask him to wait, reward him if he did, but if he popped up, we’d just ask him to lie back down, no reward. The longer he stayed, the bigger the reward. It took about two weeks. The first week showed relatively little progress, but by the second, we were able to extend his stays longer and longer between rewards, which got bigger and higher value as the intervals got longer. Eventually we arrived at the ritual we still observe. The dogs lie on their beds, a polite distance from the table until we finish dinner and are then rewarded with a “jackpot” of high value treats.
Donna in VA says
My sheltie was initially reactive towards other dogs (about 70-80% of the time there would be bark & lunge behavior – I used to count and keep a mental score each day). I initially tried “DRA” by asking him to sit and wait, watching me, if another dog passed by. I don’t think that would have succeeded, it gave him too much time to build up his anxiety while waiting for the dog to pass by. After trying some other unsuccessful deterrents that only escalated his anxiety, we settled on the “time out” of him having to lie down immediately and wait for a release if he did a bark & lunge. This was only moderately successful. I added “DRO” because I started treating and praising any behavior other than bark & lunge when we encountered another dog. Even if he was at the far end of his leash and panting nervously as we passed by, he got his treat and praise. I still do the same thing every day because I value it so highly. It was a combination of #1 and #3 in your discussion that worked, really within a very short period of time, maybe a couple of months. (Thank goodness it worked.)
LisaH says
I can’t resist sharing an early training situation when my girl was just a tiny 8-12 week old puppy. When potty training I would always go to the potty area, say “go potty” then treat her the very second she was done, which led to her fake peeing to get more treats. I wasn’t always sure if she had gone again, and again, and again, but she obviously thought the squat was what was rewarded & not the peeing! And I also find value in using both an upbeat “yes” and a more neutral “uh uh” to mean not that behavior, and my dogs get it. Sometimes we play a version of hot-cold in the yard using this …. if near the correct toy its “yes” and if moving in the wrong direction, its “uh uh”.
Heather Staas says
I’ve just adopted a new kitten, previously feral, and decided to try this approach with my German Shepherds in integrating the kitten. They are pretty good with the senior cats in the house, but that is a MUCH different experience than an 8 week old sassy kitten ripping around the place getting into everything (spinning, climbing, erratic movements, etc.). I’ve been HEAVILY rewarding just “doing nothing” in the same room as the moving kitten, gradually expecting more time and less input from me, and it is working BEAUTIFULLY. So far. Now some real interaction initiated by the kitten is starting to happen and the dogs are mostly low-arousal and responding socially appropriately without getting agitated. So far.
Sonja says
I apologize in advance for the off topic post. I know this is a behavior-oriented blog. However, it is written by a scientist with much science-minded readership.
I have a close-to-home interest in spay incontinence. One of my girls developed incontinence immediately following her spay surgery. The condition is absolutely treatable, but a certain sadness/guilt remains. A young, healthy dog developed a lifelong health condition in direct response to a decision I made on her behalf.
It’s been nearly 3 years since my girl’s surgery. We’re fine and happy. Annie only “suffers” from a daily peanut butter-coated pill. 😉 However, I want to do everything in my power to prevent this from happening to any future girly members of my “pack.” I’ve read casual implications that spaying early is a contributing factor. Annie was spayed at approximately age one year, which never struck me as particularly early.
Is there any scientific literature on spay incontinence, or is there any solid evidence that age at the time of spay is even a contributing factor? Take this as a topic request or simply a request for any information that may be floating about in one (anyone)’s head.
Wendy W says
Thanks all – this and the prior post/comments were very helpful to me a couple of days ago as I realized that I was accidently encouraging a negative behavior (defensive aggression) as a result of using a new leash.
I had clipped the new leash around my waist (the leash was designed for runners), and adjusted its length so that it wouldn’t touch the ground when my girl (Hope) was in the heel position. This adjustment made the leash considerably shorter than my standard 6 foot lead. Over the course of several days’ walks, I kept myself focused on using “body english” rather than leash corrections, and enjoyed how the hands free approach let me swing both arms and walk a bit faster.
All went well – or so I thought – until another dog lunged at Hope from the other side of a fence. Hope barked and took off, and then came to a very hard stop as she hit the end of the leash (which I couldn’t soften as I could have if the leash had been in my hand). I was surprised that she had reacted so strongly, given that she usually has no response to fence runners/barkers because I have used DRI to condition her to turn away and look for her “bark cookie” in this circumstance.
Half a mile later, she surprised me by snarling at one of her best buddies as he came up to greet her from behind his fence. Later that day, it dawned on me that the shortened leash may have been causing her to feel more vulnerable, given that she generally resorts to a dart and dash move when she wants to get away from another dog. The short leash prevented this option.
On our next walk, I used her regular 6 foot leash, and could see how much happier and relaxed she was. Thanks again to Trisha/commenters for this blog and for your ability to help me learn how to see things from Hope’s perspective!
Janice says
Don’t plow up your pasture!!!! Sorry, I know that this post was mainly about training, but I saw that and the farmer in me came out. If your pasture has perennial grasses and forbs, you’ll do it immeasurable damage if you plow it in a dry spell. That will expose the roots and all the important soil organisms to the heat and drying and then when the rains do come, there will be nothing there to hold on to the water or settle it in the soil and you will have major mud and erosion. And the exposed ground will get invaded by weed seeds and give them a toe-hold. Perennial plants know what to do in a dry spell–they go dormant. When conditions come back to their liking, they wake up and start growing again. We never get summer rain where I live (except the occasional dusting from a thunderstorm)(and I do mean a dusting, because our cars are always covered with mud splotches afterwards). Often we have no rain from July 4th until Halloween. The grasses go dormant and the animals eat it like standing hay (provided it is not over mature). But you want the soil covered with vegetation so that it doesn’t dry out more. Then the grass wakes back up when the rains come. Plowing it would be the kiss of death. Anyway, I don’t live in your climate and don’t know what it planted in your pastures but I have taken a lot of pasture management classes in recent years. You might consider spreading manure and harrowing so that when the rains return, it will kick start the grasses. Even if you have annual plants in your pasture, you don’t want to plow when it is extra dry or you will only dry the soil out further. Or another thing I have been learning about lately is see if anyone in your area is trained in permaculture and can help you with keylining your property. If climate change is really occurring, you may have more dry summers to come and will need to alter your farming techniques accordingly. Anyway, I can’t think of anything worse to do in a drought than to plow so that just burst out of me.
Trisha says
Janice, you are so dear to worry about my pasture! Thank you so much, I love that the farmer burst out… But I should have said, apologies for not doing so, that I wouldn’t consider touching it until fall, which is the best time to plant here. My intention is to do lots of research about no-till planting, if it’s appropriate, which grasses and forbs I should consider adding in if I should, etc etc. And I hear you loud and clear about climate change, it does look like it’d be wise to consider a different paradigm. Right now I am keeping the flock off the pasture, in hopes that parts of it will recover. I’m afraid some areas are just plain dirt. The combo of no rain and over 100 degree temps literally baked some of the plants into dust. The worst parts of my pasture on on top of the hill, fully exposed to the sun and convex so that water all falls away. Regrettably there is no vegetation to hold onto anything right now in those parts, but I still think I’m better off leaving it alone until I learn more. But at least there is SOME grass on the farm now, whew! (And no rain from July 4th til Halloween? Sounds like Arizona or Texas?)
Trisha says
To Wendy W: Congratulations on your successes with Hope, I’m so glad that the resources on the blog have been helpful. Keep up the good work, Hope is a lucky dog!
Janice says
Actually, I live in North Idaho. Everything in the inland Northwest has an arid climate because of the rainshadow from the Cascade Mountains–but not something I would expect someone from out of the area to know. Some areas are irrigated, but where I live there is no irrigation and so we live on what falls from the sky, usually in the form of snow and spring rain. So, here are some things for you to research: permaculture, Holistic management, keyline irrigation systems. My trek into this came with a lecture on pasture management by James Gerrish–where I realized that everything that I thought I knew about managing pastures was thoroughly wrong. So, I’ve started learned from there and discovered these other disciplines along the way that dovetail nicely. What has been so much fun about learning these new disciplines as a biologist is that I really understand their underlying concepts of the interconnectedness of nature. So that dried out hilltop you described, if you did the math, you would be astonished how many gallons of water has actually fallen there in precipitation. But it didn’t stay there because of factors of soil absorption, contour, evaporation, wind, sun….so how can you change some of those factors to keep more of the water in the soil where you need it? (I can give you one good idea right off the bat, but someone trained in holistic management or permaculture could give you 20).
One thing you will really like when you get into these disciplines is how grazing herbivores play an important role in building soil which in turn does a lot of carbon sequestration. I have always seen human, herd animals and herd dogs as an ancient trio, going back thousands of years together. But now I see them as a quartet with the soil as an additional “organism,” supported by and also supporting the other three. As a biologist, when you start learning about this to improve the drought tolerance of your own pastures, you will really be tickled by what you find. I hope to be making some major changes to my farm soon.