Questions About the LLW Video

Yesterday, I published my first tutorial to youtube.  This video features the very earliest stages of training a dog to walk politely with you.  There are a number of more advanced exercises I hope to record and share in the future, but this is how I like to start the behavior.  In the classroom, I usually start students working while their dogs are tethered, via a long leash, so that the students can focus on the mechanical skills of reinforcement delivery and return to home base, getting comfortable handling their clicker and treats.  When the handlers begin working with the leash in their hands, the dog has already learned that staying at the left hand side pays off.  It’s much easier to handle your leash, clicker, and treats when you don’t have a 95-lb dog pulling your arm off!

I’m quite new at filming and editing, so the lighting, sound, and content isn’t really perfect or refined.  Nonetheless, hopefully it gives you some ideas on how to begin this behavior with your dog.  A few viewers have posted questions or contacted me with questions regarding the video.  Here are the answers to some FAQ’s that I’ve received since posting the video:

I’ve always heard you’re supposed to use the clicker to mark the behavior you want.  Why are you clicking and treating Cuba when he is in the wrong position?

The clicker has two fundamental jobs.  It a) marks the behavior and b) announces reinforcement availability.  The former function of a clicker is its operant conditioning application, the latter is a classical conditioning application of clicker training.  Behavioral researcher Jesús Rosales-Ruiz, Ph.D. says that the “clicker is a cue to go get food.”  That is how I’m using the clicker in the early stages of loose leash walking training.  I’m classically conditioning the dog to learn that being in that spot is a predictor of good things happening for the dog.  Once the behavior has been built to where the dog is able to maintain the position and understands staying “in the pocket” is the name of the game, I like to switch to operant conditioning.  In this video, I began clicking for duration using “300 pecks.”  In other sessions, I’ll begin clicking for precision – position of the head, tightness of the heel behavior, etc.  Eventually, I will need both a true heel and a conformation gait with Cuba, both of which will be taught separately and put on cue.

What if my dog makes an error during “300 pecks”?

First, make sure you work this behavior in many short sessions.  Dogs don’t generally have very long attention spans, so quit on a success and always leave your dog wanting more.  Dogs that begin wandering off, sniffing, etc. may be telling you to shorten your session length.

In the video description, I linked to a great article from Laura VanArendonk Baugh that has a nice method for dealing with errors when using 300 pecks.  I personally change direction and resume counting the same number of steps as on the failed trial as soon as the dog catches up with me and finds the position.  If the dog makes two mistakes in a row, I need to assess the situation in more depth – is the training environment too distracting?  Should I increase the value of my reinforcement?  Should we take a break on the behavior for this session?

What if I don’t want to use a clicker?

You never have to use a clicker. There is compelling research that clicker training will allow you to train new behaviors faster and with fewer reinforcements than if you use a verbal marker but verbal markers work nonetheless.  Here are some properties which identify a good marker:

  • Is distinct and unique.  Does not sound like praise or any existing cues.  Is not heard in every day conversation.  Is said in the exact same tone, every single time.
  • Is salient – easily perceived by the dog in a wide variety of positions and environments.
  • Is consistently paired with positive reinforcement – whether or not the clicker or other marker needs to be “charged” is a topic now up for debate.  Many people simply start using the clicker to mark correct behaviors, following with reinforcement, with great success.
  • Is brief – if you are using a sound or word, think short, staccato.  Sentences don’t make great markers.

Other questions will be answered in future videos.  Thanks for your feedback everyone, hope that this blog helps clear up some of the questions you may have had!

5 comments on “Questions About the LLW Video
  1. Susan says:

    Casey,
    I thought the video was great ! You won’t please all of the people all of the time, but your very thoughtful response & reply to questions was awesome! I have always wanted to post some training videos, but haven’t had the nerve just yet, but this is truly an inspiration and I just wanted to let you know…I commend you and give it a thumbs up!
    Susan

  2. Mary Hunter says:

    Wow! He’s getting big isn’t he? 🙂
    He is so lanky, all legs, looks like he still has quite a bit of growing to do!

    Love that you are making videos now. What a great time to be making videos, since you have a young pup. I hope you’ll share more of your training sessions with Cuba.

    A couple of questions– do you find this method for errors works well with the 300 peck method?

    I sometimes follow what Alexanda Kurland teaches–if the animal doesn’t complete a certain number, reset to zero.
    Often, though, I build duration in a bit of a random, but systematic fashion. So, something like 1,1,1,2,1,2,2,2,1,2,3,2,3,1,2,3,3,3,2,3,3,3,4,3,4,2,1, etc.
    Gradually increasing the amount, and not adding the next number until the animal is pretty successful at the current one. And, constantly throwing in lower numbers that have been previously reinforced to keep our rate of reinforcement high. If the animal is making multiple errors, it usually means I’m increasing the duration too fast. It works, but I’m always looking for suggestions for more efficient ways to do it!

    I’m not sure I followed the paragraph on classical conditioning.
    The “clicker is a cue to go get food.” idea is usually discussed by JRR (and others) completely from an operant conditioning paradigm.
    The antecedent is the click, the behavior is a specific behavior or set of behaviors that the animal does to get the food, and the consequence/reinforcer is getting to eat the food.
    I really like this description of the clicker because it highlights the importance of predictable food delivery and explicitly training food delivery.
    I’ve got a few paragraphs about clicks and cues and some of the work done at UNT about halfway through this post:
    http://stalecheerios.com/blog/orca/alexandra-kurland-loopy-training/

    Mary

  3. Stephanie says:

    Thank you for the great video. I’m excited to try it.
    I’m wondering though if you are actually feeding the dog 1 treat each click? Or just letting him lick your fingers? As I don’t see you go into your bag each time to get a new treaty. Or are you holding many treats in your hands? This is where I become all stumbled; trying to get a treaty each time to reward, not dropping them on floor, being fast enough, etc.

    Thank you,
    Stephanie

  4. admin says:

    Hi Stephanie,

    I usually load my hand up with a variety of treats. I should do a video about treat delivery soon, since you’re not the only one who has asked after watching the video.

  5. Stephanie says:

    Thank you for the reply. Agree a video of treat delivery is much needed :)) . Hope to watch it soon.
    I hope you can e-mail us once you have it

    Steph

Leave a Reply

Your email address will not be published. Required fields are marked *

*