Quote: Konnie Hein

Certainly there are many ways to train an effective SAR dog, and the above is only my opinion based on my experience. I hope the info is helpful though!

You are very helpful! I have come to the realization that, as you said, there are many ways to train an effective SAR dog! I really appreciate your input as I'm taking this time while my pup is young to watch the dogs work, watch the handlers work, work on my handler skills (navigation, etc.), and learn from other handlers such as yourself and Nancy. I feel like a little SAR sponge. I actually spent all day pondering this very conversation. So I'm very open to the many different ways people address the different aspects of SAR training and please do share with me again if you ever notice I've started a thread.

Welcome to the world of SAR addiction, Jasmine!

One thing I found to be incredibly helpful was to find a good mentor with a record for repeatedly training/producing very good working SAR dogs, and then I followed that person's method and advice to the letter (perhaps you've already done this!). I asked loads of questions and soaked up all SAR-related info like a sponge too, but I always deferred to my mentor's approach. As time goes on, you'll develop your own approach and style and perhaps your mentor will learn a few things from you!

JMO... If I were you I would teach the bark alert first….then back chain it to the SAR search itself. If you back chain it, the bark alert will have the strongest reinforcement history.

Matt, can you describe how you would teach the bark alert? Thanks!

I think most SAR folks use back-chaining, but the method by which they start the alert seems to differ quite a bit. When it comes to the alert, we (my team and I) do something a little different for our USAR dogs and I wouldn't describe it as truely back-chaining. The association between the reward and the odor is actually taught before the alert behavior (although most dogs started in our training will readily bark in frustration for their reward).

Here is a blog article I wrote, explaining how we train the alert behavior:
http://focusedscenting.blogspot.com/2009/04/rewarding-with-no-bark.html

Edited by Konnie Hein (07/01/2009 10:43 PM)
Edit reason: added a sentence

 Quote: Matthew Grubb

JMO... If I were you I would teach the bark alert first….then back chain it to the SAR search itself. If you back chain it, the bark alert will have the strongest reinforcement history.

 Quote: Konnie Hein

The association between the reward and the odor is actually taught before the alert behavior (although most dogs started in our training will readily bark in frustration for their reward).

Here is a blog article I wrote, explaining how we train the alert behavior:
http://focusedscenting.blogspot.com/2009/04/rewarding-with-no-bark.html

I skimmed the article and Matt, this method seems to be what I've seen used as well...first getting the dogs to recognize a given scent is significant, then frustrating them into an alert later. But I don't know how all of the dogs were taught their alert so I may inquire about that at our next training.

I have seen backchaining used VERY successfully for recall refind alert - and it makes a lot of sense because it is a multi-step behavior, but the frustration approach and rewarding in scent seems to work real well for bark as it is very natural to build on.

First I would hold the toy above my head and elicit a bark out of the dog…. Mark it…. Then offer the toy as a reward. I would then over the course of the next session work on two or three or four…ect. barks before I would mark and reward. Then I would work on pairing the bark alert with the search command, with the continued with the exact same body positioning.

At this point we have paired the search command with the bark alert…the marker..and delivery of the toy reward. I like to change only one variable at a time in my training so I would now bring in a “victim” who will now become the vessel for the toy reward. The victim will stand in the same position I was standing and we will again elicit the bark alert for the toy… mark… reward….. and repeat and repeat.. stretching out the bark alert duration… marking and rewarding.

Again…we will only change one variable at a time…this time I will pick distance. I will increase the distance between my team and the victim. The victim is still holding the toy reward in the same manner as I held it. The command will be given, the dog released…bark alert… marker…reward.

The next variable that will be changed is that the toy reward will no longer be visible…. But the concept of marking the bark alert remains constant. Victim positioning will change as well. We want to teach the “bark alert…mark….reward” with the decoy in different positions… lying down…. Sitting… “bark alert…mark…reward”

Once the dog is solid with the bark alert on a “victim” you can then “backchain” the actual search itself. You will find that teaching the actual search itself will now be WAY easier because the dog will already have made NUMEROUS repetitions of the bark alert.

Think about how much easier SAR searches… rubble pile searches… and building searches will be when the dog has a clear understanding of what the end result will be and you have already “backchained” your bark alert.

Fun Stuff!!!

Konnie.. just looked at the blog you wrote on the focused scenting.. GREAT stuff. Operant training is really starting to gain acceptance…. The proof is in the pudding !

Thanks, Matt, for taking the time to write up such a detailed description!

And thanks for your comments regarding my blog article!