7 Common Dog Training Errors: More Cautionary Tales

I have to admit that I likely have a fair number of readers who look forward to reading about my mistakes. But hey, I asked for it, from the very first day of the blog.

My previous post on common dog training errors was very popular and I’m very happy to see it still making the rounds! So here are seven more, five of which I have personally made in spades.

(1) Too much freedom too soon 

The person who should be ashamed is me!

The one who should be ashamed about this is me!

Boy, this is an easy mistake to make. I bet a large percentage of problem behaviors and damaged property (and so-called “dog-shaming” photos) can be linked to this one simple error. Lots of times our hearts overrule our heads. Let’s say you just got a rescue dog. You feel very badly about his history. You work part-time  and you plan to crate him when you go to work. Only problem: he hates the crate. You can’t stand putting him in it the first day you go to work. The idea breaks your heart. He’s sleepy anyway when you get ready to go, so you just leave him loose in the house. You come home to poop in the corner, a chewed carpet, and some overturned plants. What a bad dog! No, he’s just a dog who hasn’t been taught the house rules yet. (Jean Donaldson’s book The Culture Clash explains this heartbreaking misunderstanding about dogs in an unforgettable way. It will change how you look at your patient, long-suffering dog forever.)

When I first got Summer, I had never had a puppy or an active adolescent dog before. I didn’t realize you couldn’t give dogs cardboard to chew on, then expect them to know not to chew up the books that were in a bookcase at floor level. I learned on the fly how to limit Summer’s opportunities to self-reinforce inappropriately, but with my two subsequent dogs I doled out freedom much more carefully from the get-go.

(This is not a how-to post, but in the case of the dog hating the crate, if you have to go somewhere before you have conditioned your dog to love a crate, most would recommend you use an exercise pen to enclose a safe space for him, or gate off the room of the house that is easiest to clear of tempting but forbidden items. And of course, leave him plenty of permissible activities, such as stuffed food toys.)

(2) Value of reinforcement too low

Last time I talked about rate of reinforcement, but what about the value? What if you ask  your dog to run a complete agility course for some kibble?  Or when she finally works up to 30 minutes quiet in the crate, you give her one piece of carrot? Or maybe you are trying not to use food at all, trying to get good results from your dog merely from praise or pats on the head (which are actually punishing for many dogs). No matter how frequently you praise, that just isn’t going to cut it with most dogs.

Chunks of dark meat chicken on a plate, round disks of dog food roll, a ziplock bag with pieces of dog food roll

Yummy stuff!

Food, especially good food, not only motivates your dog, it makes the communication in training crystal clear. When your dog gets a great treat repeatedly for the behavior you want, it makes it very clear to her that this is what pays off. There is no muddy water. If your dog is not responding eagerly in training sessions, check not only your rate of reinforcement, but the quality of it.

I have written about my own experience with Summer, who was highly distracted by her environment and just not really into the work we did. My food reinforcers, though high value, were cut in too small pieces.  Once I rectified that, the nice chunky new treats passed value to training in general and we got over the hump. She is now a training junkie and works eagerly for kibble.

I might also mention, once more, that what is yummy is defined by the dog. I was going to make a photo contrasting chunks of meat with something lower value. I thought of using bread, then remembered that my dog Summer will do anything for white bread. It pays to know these things!

(3) Over-using negative punishment

Negative punishment is defined as follows: Something is removed after a behavior, which results in the behavior happening less often. It often takes the form of a penalty or time-out.

Pulling the treat away when Zani moves out of a sit

Pulling the treat away when Zani moves out of a sit

The thing about negative punishment is that it meshes so perfectly with positive reinforcement sometimes. Too perfectly. It’s an easy default method. You start to hand the dog a cookie for staying in position. The dog starts to move out of position to get it, you pull the cookie back. You walk into the room where the puppy is in the crate. She starts to cry when she sees you. Oops! You turn on your heel and walk away. Or how about this one? You are teaching your dog the cups game, where she figure out which cup has the treat under it, then indicate that somehow. She guesses wrong and indicates the wrong cup. You immediately pull both the cups away.

These are all terrifically easy, and often effective ways to train. In all cases there is a penalty for the incorrect behavior, and it is the disappearance of the goodie the dog was  on the cusp of earning.

It surprises some people that negative punishment is at the same level on the Humane Hierarchy as extinction and negative reinforcement. Most trainers are more “OK” with negative punishment than negative reinforcement, but I think Dr. Friedman is telling us that we need to look at each case individually.

Negative punishment is punishment. It suppresses behavior. It can be unpleasant for the learner. It can directly inhibit them from trying stuff. Two of my dogs, Zani and Clara,  tend to shut down very fast if I pull an item away from them because they have taken the wrong action, as in the cups game example above.

I treat my own over-use of negative punishment as a symptom. When I find myself using it or being tempted a lot, I ask myself what it is that I have not sufficiently trained. If my dog is pulling out of position to get a cookie, there were probably holes in our stay practice. If the puppy regularly whines in the crate, I have lumped somewhere.

I’m not sure if this is an error in the same category as the others. It’s more of a value judgment. Negative punishment is still much more humane than some alternatives. But I invite you to look beyond it, whenever you find yourself or your students using it a lot.

(4) Treating in a sub-optimal position or manner

Pulling Cricket out of position for her treat

Pulling Cricket out of position for her treat

Well, there could be a whole treatise here. I am a former expert at this. You can see in this movie about Cricket in her Prime, from the very first scene, that I built into her training a little leap up for the treat in almost all behaviors. Partly because she was small, and partly because she was so intense, and entirely because I didn’t know any better. In the picture to the right, even though I had already clicked, how much better would it have been to treat her down on her mat rather than letting her jump up into the air? The position of your treat delivery can help train the behavior.

Then there’s the difference between throwing, dropping, or handing over treats. Throwing treats is very exciting and fun for lots of dogs. In certain situations it’s perfect for setting up another iteration of what you are practicing and buys you some time. So would you want to do that every time you click your dog for another increment of relaxation if that’s what you were practicing? Probably not. On the other hand, if your dog is slower than you’d like on some rapid-fire behavior, throwing treats for her to chase can amp things up.

And yes, I get the irony between the picture of my deliberately pulling Cricket out of her sit for a treat, and the picture of my pulling the treat away from Zani when she breaks position. Same picture. Hmm, I wonder how Zani learned to break position in the first place…

(5) Making training sessions look like “training” and not real life

Guilty, guilty, guilty. That’s me. This one is similar to “Failure to generalize,” in the last post but it’s more, um general. When you fail to generalize a behavior, a dog knows how to do it in one location or situation, but not another. So once your dog knows “sit” in all sorts of places and situations, is there something more you should do? You bet. Did you have your treat pouch on during all of those sessions? Or have your clicker and a pocketful of treats? Did you cut up the treats just beforehand? In other words, is everything about the situation screaming, “This is a training session?” Then good luck getting Fluffy to sit the first time your best friend comes over and you are having coffee at the kitchen table. It’s not just the possible lack of treats. It’s a completely different situation for your dog.

So first, the food. Your dog needs to learn that she might get a food treat even if she hasn’t seen all the signs of “training session.” One way to do this is to cache little covered containers of treats out of your dogs’  reach around the house and even the yard or your walk route. Casually, outside of a session, ask your dog for a sit. (Start off in the less challenging situation, of course.) Voila: out comes something really good from a jar on top of the bookcase! You can pull treats out of the sky!

Think about what else indicates to your dog that you are about to train? Do you gather up some props? Get your clicker? Put the other dog in a crate? Take your phone out of your pocket? Believe me, whatever the habits are, your dog knows them. So prepare to surprise your dog. Just like with any other training, start simple and raise your criteria. One of the main reasons most people train their dogs is to make them easier to live with. This won’t happen unless you integrate their training into real life.

(6) Clicking or marking without treating

I still see questions about this. “When can I stop treating for every click?” The answer is, “Never.” Although there are a few rarer training systems where one click does not equal one treat, if you are a beginner, forget about them for now. The clicker (or verbal marker if you use that instead) gets its power from being a perfect predictor of good things to come.

Now, it’s perfectly OK to fade the use of the clicker over time. You don’t have to click or mark every single time your dog does what you cue.  And over time, with skill, you can use food less and life rewards more. But if you click, give a treat, unless you just clicked something totally disastrous. One missed pairing out of 100 click/treats will not ruin the meaning of the clicker.* But just remember the look on your dog’s face when you don’t give them the promised treat, and do your best not to make that mistake again. Because clicking the wrong thing was your mistake, not your dog’s.

(7) Too long a delay between the behavior and the consequence: assuming the dog makes a connection when they can’t

I once read on a dog chat forum some comments by a man who was fervently defending punishing a dog when he got home and found out that the dog had done some misdeed–perhaps an elimination problem or the dog tore something up. He was incredulous that anyone would question his punishment; he said, “But dogs have great memories!”

Yes, they certainly do. His dog probably remembered peeing in the corner or how good that shoe tasted. But how exactly is the punishment supposed to be connected to that deed from hours earlier? The dog has performed hundreds of behaviors since then. Showing the dog the pee or the shoe does not connect their earlier action to whatever punishment is being doled out.

Consequences for behavior need to be very close in time to the behavior for behavior change to occur, and not just for dogs. A behavior analyst named Kennon Lattal has been the go-to guy since the 1970s for studying the effect of time delays and intervening events between behaviors and  reinforcers for people and all sorts of animals. In one famous experiment he tried for 40 days (one hour a day)  to shape a pigeon to peck a disk while delaying reinforcement for each behavior for 10 seconds. The pigeon never got there. When he changed the time delay to one second, the bird learned in 15-20 minutes.  (Paul Chance, Learning and Behavior, Fifth Edition, 2003, p. 160)

So, actually two lessons about treat timing here: when you are training, deliver those treats (or tennis balls, or whatever) as quickly and efficiently as you can. And in day to day life with your dog, don’t assume that if you give them a goodie or a talking to, that they can associate it with something they did 5 minutes or 5 hours ago.

Any of these strike home with you? Care to share? I can’t be the only one making these mistakes, can I?

This post is part of a series:

Coming Up:

  • Invisible Cues
  • How Skilled are You at Ignoring? (Extinction Part 2)
  • What if Respondent Learning Didn’t Work?

Eileenanddogs on YouTube

*There are eminent people who say you shouldn’t fail to treat even in this situation, even once.

Share Button

About eileenanddogs

Passionate amateur dog trainer, writer, and learning theory geek.

Eileen Anderson on Google+

This entry was posted in Dog body language, Dog training hints, Dogs' perceptions, Generalization, Human and dog misunderstandings, Multiple dogs, Stress Signals. Bookmark the permalink.

23 Responses to 7 Common Dog Training Errors: More Cautionary Tales

  1. Love #s 4 and 5 because I’ve made those mistakes over and over again with all kinds of different behaviors. #5 is a real challenge in multi-dog households. I feel like I make that mistake regularly when I want to work on advanced behavior or trick work and it inevitably requires separating one dog from the rest, which by default suggests a training scenario (at least when inside the house).

    And on that note, I love Charlie Bear treats for keeping in containers around the house for impromptu treating. You don’t even need to keep a lid on them which makes it that much more convenient to quickly grab and they’re cheap to buy at Trader Joe’s. 🙂

    • Thanks Emily. You are sure right about #5! I’ve been counterconditioning all my dogs to certain sounds and boy is that a challenge! Just getting everybody else out of earshot, not to mention trying to work in a “surprise” factor now and then. Thanks for the great idea about Charlie Bears. I use goldfish crackers (don’t tell the treat police) and small leftover pieces from dog food rolls.

  2. nickynockynoo says:

    I have the T shirt! The worst thing that ever hapened to me was being overconfident with LLW. I had the lead looped through the middle finger of my right hand, a few treats on top and the clicker over them, leaving my LH to deliver treats (Thinking to myself how negative some people can be about needing 3 hands to clicker train) All went fine until George saw a cat………………..yes, he pulled and my thumb clicked!

    Re formal training sessions “But if you click, give a treat, unless you just clicked something totally disastrous. One missed pairing out of 100 click/treats will not ruin the meaning of the clicker.*
    One of the best things I learned at an Emily Larlham seminar was; If you click at the wrong time, move the dog to another position and carry on as dogs do not generalise very well. That tip really boosted my confidence.

    Love this blog and the way you post so often. Thanks Eileen.

    • Nice tip from Emily! Wow, that was a pretty bad click there. I’ve been working on the hold for retrieve and it is so devilishly hard not to mark the release instead of the hold as we are increasing duration. But I treat all those misses. Goes with the territory. It’s making me get my act together. Thanks so much, Nicola.

  3. Pingback: 8 Common Dog Training Errors: Cautionary Tales | eileenanddogs

  4. Reblogged this on BFTG.

  5. mrsskeats says:

    I love this. I had both my former dogs from puppies and instinctively they learned what I wanted, because it was all they knew – and we somehow muddled together. Many folk would say how well trained my dogs were. I simply am quite an arse basically about dogs and rules. Not on the bed, floors are for dogs, not on the furniture (same), eat when I say ‘ok’ (that’s so if as a pup they thought picking up that old mouldy sweet of off the pavement was good I could stop them with a command). So when I got Annie, a romanian rescue aged one, who had been street living then kenneled I started the same way – and it went wrong sometimes. I agree you can’t give a dog an old shoe and expected them to know new shoes are no go etc. But to train her was difficult. She wouldn’t do it for food. I believe she associated food with being captured, and would baulk at the idea. So, I read up …lots… shouting is a definite no no. A low growl command when she’s been naughty is far more effective but only effective when caught IN the act, and not hours later as per your comment in 7. I find also when she comes for fuss, if she’s been naughty and knows it, she lowers her head and ears. I simply turn my back on her and she goes to her bed. Then after a short period I would show her the chewed letter or blanket, crouch down in front of her and simply say (not shout) “Mummy is not happy with you right now, this is not good.” Then I’d walk away taking damaged blanket or whatever with me and leave her. Effectively giving her time to realise that the behaviour meant nice things weren’t coming her way. On the same note, when she did go a few hours without chewing her blanket etc, i would simply not remove it. She soon learned chew- plastic bed, no chew = comfy bed. And so on. One guy wrote in his paper that we may as well talk russion to our dogs (unless the dog is russian then talk chinese. It’s our posture and body language that tells them how we feel toward them. So I work hard on that, the talking is just my way of telling me what I’m telling her. She’s 2 now and a good dog – mostly.

    • It’s amazing what dogs teach us, isn’t it? I too had had “easy” dogs until Summer came along. I also had the experience of a dog who thought food was a trap (Clara the feral dog) but she got over it fast since she was a puppy. If I may suggest, don’t give up on using food to train your dog though. It can make things so crystal clear, and they find the training so fun.

      It sounds like you read a lot. You might want to check into this article where they did an experiment to see whether dogs’ acting “guilty” really meant that they knew they had done something. It turns out that dogs act like that to appease cross or angry humans, and there is no connection between naughtiness and acting guilty. Just some food for thought.

      Great point about verbal language vs our body language. Just about every time I start to think my dogs really know what an English word means, I realize I am still doing something with my body that they are really paying attention to! Thanks for the comments!

  6. Jean Becker Inman says:

    I’m sure I am guilty of all of these from time to time. Some more than others. But overall we are making progress despite me being a less than great handler.
    I do have treat containers all of the house and in the car. My personal dog (I foster as well) has learned that if I am wearing something with pockets, there are probably treats in those pockets..LOL.

    • Dogs learn those signals, don’t they? I almost always have treats in my pockets, too. Sounds like you are doing well!

    • Chloe Colleen says:

      A few cookies tucked in your bra can help make for an effective surprise reinforcement. Socks and hats even, nothing’s off limits!
      (Also, every word I tried to replace “cookies” with was also a euphemism for breasts, and adding descriptors just made it worse–for example: dry biscuits. Sorry…) .

      • Good job on the cookie terminology (and suggestion), Chloe! Just yesterday I tucked an actual cookie into my back pocket. My dog was too distracted at the time, I suspect, to smell it. She was pleasantly surprised!

  7. nickynockynoo says:

    “I’ve been working on the hold for retrieve and it is so devilishly hard not to mark the release instead of the hold as we are increasing duration.”

    I tried for months to back chain the retrieve. George would fetch balls but was not interested in dumbbells, hose pipe or dowling as a hold object. I used Jean Donaldson’s “Shaping the Dead Retrieve” from her book, “How to Train Your Dog Like a Pro” The hold just fell into place easily afterwards (as a separate training session) A lovely quote from that chapter, “My goal is for you to focus on razor-sharp timing, sophisticated criteria setting and feeding for position!”
    I don’t think this would mess up Levels training either as they are 2 behaviours, until joined up of course.

    • I need to get that book! I have actually gotten over the hump with the hold. We are at 3-4 seconds now and going strong (cue cheering!).

      I once saw a great video by Jean Donaldson that she accidentally set as public on YouTube for a short time, where she was teaching a dog to pick something up. She had a brilliant system for preventing clicking the release. I bet it’s similar to the book. Thanks for the great reference.

      • Rita says:

        I’m just wondering if having conversations or run on sentences with a dog is maybe as harmful as it is pointless? I know that they pick up on certain words and commands. Could they be getting confused by too many words at once?

  8. Pingback: Service Dog Training Rancho Bernardo CA Dog Training | PRO-TRAIN - Best Dog Training

  9. Rita says:

    For example: my boyfriend will say something to the effect of “sammy if you don’t CALM down we’re not going for a WALK” i feel that my dog then only maybe picks up on the “calm” +\ “walk”. In your opinion is he confusing her. I tell him he’s ineffective w her because she doesn’t understand all his talk, is this true?

    • Rita, I think you are making a great point. Although dogs can grasp so much of what we mean from our tone, body language, and actions, I think most people over-estimate that exactitude of dogs’ grasp of human language. One example of that is the practice of giving a cue: “Sit!” then following it with praise that includes the cue again: “Good sit!” The better the dog understands a cue, the less sense the praise is going to make since it repeats the cue. We expect them to parse the difference between a cue and a modified noun (modified with the adjective “good”). If we have successfully taught them that “Sit” is a cue, their natural response would be to try to sit again, and indeed some dogs will grind their butts into the ground visibly if told to sit when already sitting. It’s more sensible with most dogs just to say “Good!” or “Good girl!” or another phrase that we have taught them means they are doing a good job.

      As for your boyfriend’s statement: Sammy probably doesn’t understand all those words. But she may understand his tone to mean: Shape up or things are not going to go your way. But remember, dogs do what works. If Sammy regularly gets taken for a walk shortly after jumping around and acting crazy, that behavior is being reinforced bigtime. So whatever your boyfriend is saying is less important than the fact that the walk usually happens, whatever her behavior. He may suppress her exuberance momentarily with his statement, but if she acts that way every time they get ready for a walk, the behavior is being reinforced.

      Hope I didn’t go too far off on a tangent here. I’ll be doing more of these posts in the future, and you’ve given me an idea to use (I’ll give you credit). Thanks for writing!

Comments are closed.