Dog Forum banner
Status
Not open for further replies.
1 - 7 of 7 Posts

·
Registered
Joined
·
9 Posts
Discussion Starter · #1 ·
Hello everyone! My name is Gia. I live in Costa Rica and am working towards becoming a knowledgeable, creative, and competent dog trainer. Unfortunately, the resources in my country are limited and I have yet to find any guidance here. I educate myself as best I can through researching online, reading books, watching videos, taking online courses, and building up my hands-on-experience with animal training and behavior modification. But sometimes I feel very alone and overwhelmed. Though I usually tend to just "lurk" on forums and not participate, I wanted to try to step out of my comfort zone and engage with all of you kind folks.

I use a clicker or marker word, and I could really use some help in regards to training certain cues and how best to mark the behavior. I don't want to be sloppy or careless in my clicking; I realize that timing is very important, and I want to keep it as clear and crisp as possible.

I really enjoy the work done by Emily Larlham (Kikopup), and I think she's an amazing trainer. I like the way that she teaches stay with the stay behavior being part of the sit or down position (basically, the dog sit or lies down and remains in that position until given the release cue), versus teaching stay as a separate behavior.

However, I noticed that in the comment section of the sit/stay video, many people were confused about the part where Emily says to click and treat as fast as you can while the dog stays in position, since the click typically signals the end of the behavior (that the behavior was completed). My understanding is that Emily is clicking/treating so quickly so that the dog is reinforced for the sit and the clicks/treats are coming so rapidly that the dog doesn't have time to get up. That being said, I have wondered if it might be more clear for us humans and dogs if, at this point in the exercise, we do one click and then the rapid-fire treat delivery (without further clicking before each treat).

I have a post on my blog about teaching the stay "Kikopup style” (which includes the original video, Emily’s instructions, and some of my thoughts), and I made an edit at one part to mention that I felt it may be clearer for the dog if we only clicked once and then did the rapid-fire treat delivery. Then I was recently looking over all of Emily's instructions that I wrote out, and I realized that there were other places that I should possibly modify to go along with the first edit. For example, Emily talks about proofing the release cue and clicking as the distractions occur. However, despite the click, the dog is supposed to stay in position until given the release cue. What about simply praising the dog and treating, but not clicking until the release cue is given?

- - -

Then I've also been thinking about other cues... for example, the way I've learned to teach a "leave it" is to click and treat for the dog staying away from the food/object(s). You don't want to create a behavior chain of them trying to mug your hand and then backing away, so you continue to click and treat for short time periods of them ignoring the food/object(s). I feel like in this exercise it works because the dog doesn't have to stay in one position or do exactly the same thing, as long as they are ignoring/turning away from/moving further away from the food. But according to my earlier statements, should I just be clicking once and then continue rewarding for the behavior of staying away, without further clicking?

What about teaching a dog to enjoy their crate? In one of Emily's videos, I saw her clicking and treating multiple times for a dog remaining inside the crate. Would it be better to click once and then offer various [eventually more spaced out rewards] for the dog remaining inside of the crate, though the rewards aren’t paired with further clicks?

*
In my mind, the examples I just gave of the "leave it" cue and the crate differ from the sit/stay cue because the dog is able to move around and sort of begin the behavior again (if that makes sense), whereas with a sit/stay the dog is supposed to continuously hold the position (and during Emily's example of the sit/stay, the clicking does not mean that the dog can move). For example, if I'm clicking and treating for a dog staying in the crate, the dog is still allowed to change positions, move around, do as they'd like--what they're being rewarded for is the act of being inside of the crate. *

I would be extremely grateful for any help that you can give me. Thank you so much.

- - -

After writing all of that out... I just found this article -- "Does the Click End the Behavior?" from Karen Pryor's website. So if I were to apply the information from that article... Let's give the example of creating pleasant associations with being inside the crate. You would start by clicking the dog for entering the crate. One click and treat. As time went on, the dog would enter, and you would wait longer and longer before clicking... so only one click and treat for the dog being inside the crate (the click/treat would just take longer to be delivered). This is different than what I have seen of Emily's training.

Helpful excerpt from the article above:

Can the click carry two meanings?: 1) yes, that is the behavior I want, and 2) please continue doing the behavior. Can it be both an event marker and a keep-going signal? Karen responds: "If you switch the click from being a marker signal to being a kind of encouragement or 'keep going' signal, you are going to be in trouble. When you give random clicks during a continuing behavior (while training a "stay" for example), unpaired with any other event, you separate the click from the reinforcer." Studies at the University of North Texas have demonstrated that to maintain the strength of the association of the click with the reinforcer, you need to keep those two items paired and associated in real time as much as possible, or the power of the click will be reduced until it is meaningless to the animal.

Sometimes, people click and also treat repeatedly during a long-duration behavior, but don't let the behavior stop. In this case the click continues to be paired with the reinforcer, but the click actually has no information in it at all, since it is not marking a clearly identifiable behavior. You have eliminated its main purpose, as a marker signal.

So, if the click must be maintained as an event marker, rather than a keep-going signal, and we don't want to click until the behavior ends, how do we build enduring behaviors?

... Mostly people fall into the trap of thinking they have to click "during" a behavior, once or many times, because they don't understand that the way to build a long-duration behavior is to gradually increase the length of time you are asking for; and click only at the end of that length of time.

 

·
Registered
Joined
·
316 Posts
Good questions!

I'd say that a click ends the behavior, BUT, ending the behavior doesn't have to mean breaking position. "Sit" doesn't have to end with the dog standing up; "sit" can simply end after two seconds (the behavior is "hold sit position for two seconds," so after two seconds, it ends, even though the dog remains sitting). By clicking and reinforcing her dog in position, without offering opportunities for the dog to break position, Kikopup is making sure that the click does not inadvertently become a cue for the dog to break position (unlike her "release" cue, which does signal to the dog that breaking position will earn reinforcement).

A sit-stay seems like one long duration behavior, but Kikopup is treating it like a behavior chain. That means she's reinforcing multiple behaviors in sequence (sit, hold sit position, hold sit position, hold sit position, change position/release on cue). Each behavior is clicked separately during the initial learning stage, and criteria is raised later. By making sure that her click does not signal "break position and run to me for a treat," Kikopup is setting herself up to be able to teach future behavior chains without having her dog repeatedly break position after each click. For instance, in crate training, a click might mark a specific relaxation behavior (I use it this way, during biofeedback exercises), without also meaning "running out of the crate will earn you a cookie." But my crate release cue, "break," does mean "running out of the crate will earn you a cookie." Two different signals!

I would also argue that there is little meaningful difference between click/treating and just treating in this situation. Both events mark behavior, it's just that the timing may be slightly different (it's easy to get that razor-sharp precision with a clicker, but not that much harder to deliver a treat with outstanding timing if you happen to already be standing right by your dog's head!). Feeding a treat shouldn't be a cue for the dog to break position. Neither should a click. So maybe cutting out clicks would make it easier for some trainers, but it shouldn't really make a difference to the dog.

Where I do think problems arise are with clicks that don't immediately pair with reinforcement (for instance, trainers who reinforce a duration behavior by going "click, click, click, treat"), or clicks that are inconsistently timed to mark specific behaviors (clicking so fast that the dog can't even think about what's going on, for instance, or spacing out clicks so slowly that the dog starts to think that "hold sit position" isn't working anymore and is motivated to try something different). Kikopup times her clicks so that there is room to think between them, and also does an extensive "proofing" session aimed directly at teaching her dog that the identified behavior is "hold sit position" even in the face of escalating distractions. By consistently using the click to mark instances where her dog chooses to hold a sit position, she's marking and strengthening that behavior, which could theoretically build to a pretty bomb-proof sit-stay.

Hope that helps a bit and wasn't too convoluted. Thanks for the opportunity to think that through, which probably helps improve my own training too!
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #3 ·
Good questions!

I'd say that a click ends the behavior, BUT, ending the behavior doesn't have to mean breaking position. "Sit" doesn't have to end with the dog standing up; "sit" can simply end after two seconds (the behavior is "hold sit position for two seconds," so after two seconds, it ends, even though the dog remains sitting). By clicking and reinforcing her dog in position, without offering opportunities for the dog to break position, Kikopup is making sure that the click does not inadvertently become a cue for the dog to break position (unlike her "release" cue, which does signal to the dog that breaking position will earn reinforcement).

A sit-stay seems like one long duration behavior, but Kikopup is treating it like a behavior chain. That means she's reinforcing multiple behaviors in sequence (sit, hold sit position, hold sit position, hold sit position, change position/release on cue). Each behavior is clicked separately during the initial learning stage, and criteria is raised later. By making sure that her click does not signal "break position and run to me for a treat," Kikopup is setting herself up to be able to teach future behavior chains without having her dog repeatedly break position after each click. For instance, in crate training, a click might mark a specific relaxation behavior (I use it this way, during biofeedback exercises), without also meaning "running out of the crate will earn you a cookie." But my crate release cue, "break," does mean "running out of the crate will earn you a cookie." Two different signals!

I would also argue that there is little meaningful difference between click/treating and just treating in this situation. Both events mark behavior, it's just that the timing may be slightly different (it's easy to get that razor-sharp precision with a clicker, but not that much harder to deliver a treat with outstanding timing if you happen to already be standing right by your dog's head!). Feeding a treat shouldn't be a cue for the dog to break position. Neither should a click. So maybe cutting out clicks would make it easier for some trainers, but it shouldn't really make a difference to the dog.

Where I do think problems arise are with clicks that don't immediately pair with reinforcement (for instance, trainers who reinforce a duration behavior by going "click, click, click, treat"), or clicks that are inconsistently timed to mark specific behaviors (clicking so fast that the dog can't even think about what's going on, for instance, or spacing out clicks so slowly that the dog starts to think that "hold sit position" isn't working anymore and is motivated to try something different). Kikopup times her clicks so that there is room to think between them, and also does an extensive "proofing" session aimed directly at teaching her dog that the identified behavior is "hold sit position" even in the face of escalating distractions. By consistently using the click to mark instances where her dog chooses to hold a sit position, she's marking and strengthening that behavior, which could theoretically build to a pretty bomb-proof sit-stay.

Hope that helps a bit and wasn't too convoluted. Thanks for the opportunity to think that through, which probably helps improve my own training too!
Thank you so much for your detailed response! That was excellent. I really appreciate you taking the time to help walk me through this. (Also, my mom was very impressed by you and she said that your reply was very well thought out.)

If you have time for further questions... How would you explain the concept of clicker training in a clear, simple way to a beginner? Should we just be saying that the click marks the behavior as correct and means that a reward is coming? Am I over-complicating this by thinking so much about whether or not the click ends the behavior (and how relevant is it, really)? What are your thoughts on the article from Karen's website? I do agree with you that each click should be paired with a treat and that we should be careful with the timing of the clicks so as not to overwhelm or confuse the dog.

Thank you very much for being so kind and helpful; I want you to know that it really means a lot to me and has truly helped me to think through all of this better. I also wanted to ask you if I could quote from your explanation for my blog post. If you would be comfortable with that, please let me know how I should quote you and if you would like me to link to a profile/page of yours. Thank you so much once again.
 

·
Registered
Joined
·
316 Posts
My thanks for the kind words from you and your mom! You're very welcome to quote the above (you can attribute it to my screen-name, "SnackRat," if you like, or to "a poster on dogforum.com," but no other link is necessary!), and I appreciate you asking.

I usually describe a clicker as a way to tell a dog, "Yes, THAT, do THAT again!" It marks behavior, and then we follow that behavior with a reward so that the dog wants to do the marked behavior again. That's a simplification, of course, but is enough for most people to get started. Once people start to shift their focus to saying "yes" to their dogs, instead of "no," great things can happen!

The theories behind training interests me, so I usually think it's interesting to consider questions like "does a click end the behavior" and "what does "ending the behavior" look like?" If you find those questions interesting, by all means pursue them...but be prepared to change your mind a few times as you keep learning, talk to other trainers, or find a situation that totally challenges what you thought you knew before!

I don't particularly disagree with anything in the Karen Pryor Clicker Training | The Leader in Positive Reinforcement Training article. I think a click "ends" a behavior in that it marks a successful performance, but that doesn't have to mean that the dog immediately performs a totally different behavior. Behaviors don't happen in isolation -- first one behavior, pause, then another behavior, pause, then another behavior -- but are always continuous and overlapping. So when we click, we are always usually reinforcing many behaviors in a continuum, which means that when I am training a dog, I have to consider what I want it to do after I click as well as before. Do I want the dog to jump up and run around in circles? Then I should reinforce that! Do I want the dog to maintain position until a verbal release? Then I should reinforce that! It's all just an ongoing back-and-forth form of communication, with many different opportunities to shape the dog's behavior. Lots of opportunities to make mistakes too, of course, but how else would we learn?!
 

·
Banned
Joined
·
1,398 Posts
- - -

Then I've also been thinking about other cues... for example, the way I've learned to teach a "leave it" is to click and treat for the dog staying away from the food/object(s). You don't want to create a behavior chain of them trying to mug your hand and then backing away, so you continue to click and treat for short time periods of them ignoring the food/object(s). I feel like in this exercise it works because the dog doesn't have to stay in one position or do exactly the same thing, as long as they are ignoring/turning away from/moving further away from the food. But according to my earlier statements, should I just be clicking once and then continue rewarding for the behavior of staying away, without further clicking?
Welcome Gia,

Sounds like you have quite a bit of passion for dog training, that's great!

I have never truly clicker trained any of my dogs as I feel my verbal marker can be altered to impart additional impact when the moment is proper.

I was always under the impression that once a dog is trained to sit, down, or stand, the stay or wait is implied. However, when teaching a pup a sit, down or stand in the beginning, one has to break it up into at least two segments. The first segment obviously being the actual sit, down or stand and then is linked to a wait or stay command once the position is mastered. Ultimately, a sit, down or stand command includes the dog remaining in position until released, so the wait or stay is implied. To me, a wait means the dog will come to me when I recall the dog whereas a stay means the dog will stay on its mark until I return to the dog's side and release it or continue on. Of course there are times where the "wait" might not have the dog return to me but move to an area I direct the dog to go.

I would have the same opinion as SnackRat that the marker would signify to the dog that the dog has accomplished the task properly, no more or no less.

The reason I chose the portion of your original post above is the "leave it" command and training has always been a pet peeve of mine. All too many people use food or objects in their "leave it" training that the dog will ultimately possess when the handler allows. To me that is a form of impulse training coupled with a wait/stay command. The "leave it" command is reserved for items which the dog will never be allowed to gain possession of because of potential harm to the dog. I trained the "leave it" with electrical cords, anti-freeze, gasoline, shoes, clothing, counter tops, table tops, mushrooms, dangerous plants, TV remotes etc. I am probably wrong but the goal of leave it training is to train the dog, when it is not being monitored by its human, that the dog will leave the items alone which it has been trained to leave alone. It's wonderful training to get your dog educated properly so it can remain in the house without any restrictions or crating.

Anyway, a click or verbal marker properly timed makes a huge difference in a pup's education and if the consistency of the marker remains, it makes it easier for a pup going forward as to what it is trying to accomplish as more advanced training takes place. I equate my verbal marker to the click of a clicker and when my pups have nailed something I use my verbal marker with more enthusiasm but rarely if ever repeat my verbal marker. The dog either accomplished the behavior or failed, so repeated verbal markers or clicks in my estimation might confuse the dog because the consistency of the marker has been changed.
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #6 ·
My thanks for the kind words from you and your mom! You're very welcome to quote the above (you can attribute it to my screen-name, "SnackRat," if you like, or to "a poster on dogforum.com," but no other link is necessary!), and I appreciate you asking.

I usually describe a clicker as a way to tell a dog, "Yes, THAT, do THAT again!" It marks behavior, and then we follow that behavior with a reward so that the dog wants to do the marked behavior again. That's a simplification, of course, but is enough for most people to get started. Once people start to shift their focus to saying "yes" to their dogs, instead of "no," great things can happen!

The theories behind training interests me, so I usually think it's interesting to consider questions like "does a click end the behavior" and "what does "ending the behavior" look like?" If you find those questions interesting, by all means pursue them...but be prepared to change your mind a few times as you keep learning, talk to other trainers, or find a situation that totally challenges what you thought you knew before!

I don't particularly disagree with anything in the Karen Pryor Clicker Training | The Leader in Positive Reinforcement Training article. I think a click "ends" a behavior in that it marks a successful performance, but that doesn't have to mean that the dog immediately performs a totally different behavior. Behaviors don't happen in isolation -- first one behavior, pause, then another behavior, pause, then another behavior -- but are always continuous and overlapping. So when we click, we are always usually reinforcing many behaviors in a continuum, which means that when I am training a dog, I have to consider what I want it to do after I click as well as before. Do I want the dog to jump up and run around in circles? Then I should reinforce that! Do I want the dog to maintain position until a verbal release? Then I should reinforce that! It's all just an ongoing back-and-forth form of communication, with many different opportunities to shape the dog's behavior. Lots of opportunities to make mistakes too, of course, but how else would we learn?!
Thank you so much. I will probably write something like "from SnackRat, a member on Dogforum.com."

I agree that when it comes to the theories of training, there's definitely a lot of space for creativity and different viewpoints. As I mentioned earlier, I'm trying my best to work toward becoming a knowledgeable "professional" here in Costa Rica and educate others in regards to training & behavior modification. This country definitely needs some help with dog culture, animal care standards in general, and implementing kind training methods. I want to take action in encouraging movement towards these goals, and I don't currently have as much guidance or support as I wish I did. So I do feel a lot of pressure to somehow "get it right" and know all the answers. I really liked what you said about "Lots of opportunities to make mistakes too, of course, but how else would we learn?!" and I think that's the part that I'm struggling with--that I have so much to learn, but maybe not the best resources here, and how do I effectively and responsibly help others in the meantime?

How do you personally teach a sit/stay, if you don't mind me asking? Thank you for sharing your thoughts and spending your time to help educate me.

Welcome Gia,

Sounds like you have quite a bit of passion for dog training, that's great!

I have never truly clicker trained any of my dogs as I feel my verbal marker can be altered to impart additional impact when the moment is proper.

I was always under the impression that once a dog is trained to sit, down, or stand, the stay or wait is implied. However, when teaching a pup a sit, down or stand in the beginning, one has to break it up into at least two segments. The first segment obviously being the actual sit, down or stand and then is linked to a wait or stay command once the position is mastered. Ultimately, a sit, down or stand command includes the dog remaining in position until released, so the wait or stay is implied. To me, a wait means the dog will come to me when I recall the dog whereas a stay means the dog will stay on its mark until I return to the dog's side and release it or continue on. Of course there are times where the "wait" might not have the dog return to me but move to an area I direct the dog to go.

I would have the same opinion as SnackRat that the marker would signify to the dog that the dog has accomplished the task properly, no more or no less.

The reason I chose the portion of your original post above is the "leave it" command and training has always been a pet peeve of mine. All too many people use food or objects in their "leave it" training that the dog will ultimately possess when the handler allows. To me that is a form of impulse training coupled with a wait/stay command. The "leave it" command is reserved for items which the dog will never be allowed to gain possession of because of potential harm to the dog. I trained the "leave it" with electrical cords, anti-freeze, gasoline, shoes, clothing, counter tops, table tops, mushrooms, dangerous plants, TV remotes etc. I am probably wrong but the goal of leave it training is to train the dog, when it is not being monitored by its human, that the dog will leave the items alone which it has been trained to leave alone. It's wonderful training to get your dog educated properly so it can remain in the house without any restrictions or crating.

Anyway, a click or verbal marker properly timed makes a huge difference in a pup's education and if the consistency of the marker remains, it makes it easier for a pup going forward as to what it is trying to accomplish as more advanced training takes place. I equate my verbal marker to the click of a clicker and when my pups have nailed something I use my verbal marker with more enthusiasm but rarely if ever repeat my verbal marker. The dog either accomplished the behavior or failed, so repeated verbal markers or clicks in my estimation might confuse the dog because the consistency of the marker has been changed.
Thank you very much for your kind welcome.

Could you tell me more about how you personally teach a sit/stay? From what I understood, you feel that it would be more clear to only click/use a marker word once and then give various treats (and continue praising?) rather than incorporating additional clicks at that point in the exercise. I was also concerned that "repeated verbal markers or clicks... might confuse the dog because the consistency of the marker has been changed."

I really appreciate all your feedback in regards to teaching "leave it." When I teach "leave it" [with food], I've been working on giving the dog a treat reward from somewhere else and not allowing them access to what's on the floor. I was hoping that this would make it more clear for the dog since, as you correctly mentioned, the "leave it" cue will often be used with "items which the dog will never be allowed to gain possession of because of potential harm to the dog." I think that all the "leave it" training that you did is wonderful and practical.

When it comes to training a "leave it," I know dogs that wouldn't be interested in trying to grab the electrical cords or other items that you mentioned. (Of course, I do realize that when they're alone and bored, that may be another story.) So I do see training "leave it" with food as being useful in the sense that they might start understanding that even if they want something, they need to leave it alone, and then we could apply that to the other objects that you mentioned (which, while not being so exciting to them while you're there, might be more interesting when they are by themselves and restless.) I hope I expressed this properly, and I'd love to hear more about what you think.
 

·
Banned
Joined
·
1,398 Posts
Thank you very much for your kind welcome.

Could you tell me more about how you personally teach a sit/stay? From what I understood, you feel that it would be more clear to only click/use a marker word once and then give various treats (and continue praising?) rather than incorporating additional clicks at that point in the exercise. I was also concerned that "repeated verbal markers or clicks... might confuse the dog because the consistency of the marker has been changed."

I really appreciate all your feedback in regards to teaching "leave it." When I teach "leave it" [with food], I've been working on giving the dog a treat reward from somewhere else and not allowing them access to what's on the floor. I was hoping that this would make it more clear for the dog since, as you correctly mentioned, the "leave it" cue will often be used with "items which the dog will never be allowed to gain possession of because of potential harm to the dog." I think that all the "leave it" training that you did is wonderful and practical.

When it comes to training a "leave it," I know dogs that wouldn't be interested in trying to grab the electrical cords or other items that you mentioned. (Of course, I do realize that when they're alone and bored, that may be another story.) So I do see training "leave it" with food as being useful in the sense that they might start understanding that even if they want something, they need to leave it alone, and then we could apply that to the other objects that you mentioned (which, while not being so exciting to them while you're there, might be more interesting when they are by themselves and restless.) I hope I expressed this properly, and I'd love to hear more about what you think.
There are numerous items which a dog will find "interesting" which are not food related such as shoes, clothing, counter tops ( perhaps food related at times ), sprinklers out in the yard, garbage cans ( perhaps food related ), furniture, remote controls, most anything a dog can get its mouth on which would be either harmful to the dog and/or destructive if chewed. My approach is fairly simple; using a lure such as a piece of food ( leave it item) which the dog might ultimately consume whether it's the same exact piece or not will create confusion for the dog. For instance, a few of my dogs when they were pups would eat their own crap like many dogs do when they are young, so I immediately took the position of using their poops as a "leave it" item. The reward for leaving any of these items alone was a food reward in the beginning along with praise and then faded out the food reward. Most all pups are notorious chewers as you probably well know so the food reward appeals to their food drive and the pup begins to make the connection and is offered a choice of sorts. Once the pup begins to make the proper choice on its own, shaped by conditioning, the dots are more easily connected and this is why I do not use a food scrap as the lure for leave it training and only as the reward. However, I have used a food lure numerous times in teaching the dog impulse control at a basic level. The dog waits for the release to gain possession of a food scrap or toy but this lesson the dog is learning is simply it must wait for a release command when commanded as such, nothing more nothing less. The significant difference is the dog awaits a release to gain possession and focuses on the behavior required ( a wait essentially ) while "leave it" training tries to ingrain the mentality of never taking possession, much like teaching bite inhibition for most dogs.

Thinking about your multiple clicks during an extended sit/down/stand. Too many people when teaching a dog to stay in a particular position will use the command numerous times because they are expecting the dog to fail as the duration is increased. Maybe one would be wiser, to give the marker immediately once the dog's butt hits the ground in a sit and then after a very short duration give one more marker and then release the dog and reward appropriately. Increase the duration systematically and take advantage of the dog's anticipation to hear the ensuing marker as the duration increases. Eventually as the duration increases to whatever length you desire the dog will learn it must stay or wait in a particular position until you release via your marker of choice. The simpler one can make it for the dog to comprehend the desired behavior, the better the dog will perform.

Chances are the dog will break in the beginning and this is when you somewhat need to reset the dog and start anew but make certain that you set the dog up to succeed by not making the holds in position too long in the beginning. It's all about patience and understand that a new behavior being taught especially to a pup can be confusing, so simpler is better.
 
1 - 7 of 7 Posts
Status
Not open for further replies.
Top