Thursday, April 24, 2014

Is Your K9 Nose Work Dog Working You?

Here is a link to a study done with the goal of testing the influence of handler beliefs on dog detection outcomes: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3078300/


In case you aren't compelled to comb through these studies, I'll share with you the major differences in their findings: the first study found that handler beliefs influence dog detection outcomes in a way that produces false alerts (a handler interpreting dog's behavior or trained response as a positive sign of the presence of a trained scent), the second study found that "the dogs indicated the target scents independently of handler beliefs and decoy scents." The second study noted that handler beliefs do seem to increase false alerts, but not in a statistically significant way.

I'm not taking a position on the topic of the studies in this post, rather, I want to spark some thinking on the part of nose work handlers as to how we interact with our dogs in searches, how we get to know our dogs, and how we learn to trust our dogs and be great teammates. 



This happened the other night in one of my classes during a practice run with 10 white ORT-style boxes, one containing a target scent. The location of the odor box was known to the handler.

Handler and dog stand at the start of the search.

Handler: Which box is it in?
Me: Third box on the left.

The team begins the search. The dog, an earnest, hard-working black lab, works boxes and the environment in a search for odor. As they work, the handler ends up turned around and facing the start line, watching her dog work back towards the start of the search. There's a strong head turn as he passes by the odor box, followed by a head drop and intense sniffing of the odor box, punctuated by continued sniffing of the odor box as the lab pushes it across the floor.

Me: (gentle urgency) Reward. Uh, get in and reward!

Handler dutifully - yet confusedly - rewards dog.

Handler: That's it?! I thought you said the third one on the left (points to third box on right)?

In this scenario, the dog actually found the odor despite the handler's belief that odor was not present in the box the dog was indicating. This was not a double blind study, not a blind search, just an example of a good dog doing his job despite unintentional handler interference.

What I like about this example is how well it worked out for the dog in terms of being able to focus on his task and show clear behavior indicating the source odor box. Without intending to do so, the handler turned down the background noise for the dog as he searched, allowing him to search, find, and be rewarded. When I say noise, I mean the various cues that a handler gives - sometimes subconsciously - during a search.



When thinking about handler beliefs influencing the dog, don't forget to think about the training the dog is receiving, and how that is affecting the outcome of the search. Well-timed reward for the dog finding source odor is a major factor affecting the outcome of the search. If your training is leading the dog to believe his reward will come for something other than finding source odor, then you've got a problem that won't necessarily get better with improved handling.

If your training is sound, and the dog is getting rewarded for being at source, then handler beliefs, and handler cues, could be holding up the success of the team. For our purposes, we'll consider handler beliefs to cause most handler cues. For example, thinking odor is or is not present in an area based upon factors other than your dog's behavior (the classic belief, "they would never hide it in the trash can") is a handling belief that may change your body language or facial expressions, producing a cue that your dog reads that interferes with his ability to find source odor.

Here are some handler cues that can make it difficult for the dog to be clear in his task:

- hand in the treat pouch in the moments leading to your dog finding source odor
- stopping and facing the odor source at the slightest behavior change from your dog
- guiding your dog to an odor source other than as part of a specific training exercise
- moving quickly and in a disinterested way through parts of a search area you know contain no source  odor to get to parts of the area you know are productive
- verbally marking your dog's behavior at the odor source
- rewarding your dog away from source after he's found source odor and given an indication behavior

Here's why these can be challenging cues for the dog: these cues almost never occur in training when odor is not present, and most of these cues typically precede a reward delivery. In the case of moving through known unproductive parts of a search area, the reverse is true, the handler almost never acts as if odor is present in those areas, but almost always leads the dog to a productive part of the search in a more direct manner. As handlers, we are being studied by our dogs just as much as they are being studied by us, only, our dogs are much more skilled at picking up on our behavior and giving us what we're asking for - whether we're aware it's what we're asking for or not! 



What can you do to minimize handler influence on the dog's indication of source odor? Here are some biggies in my opinion:

Spend more time building independence in your dog through self-reward searches - this is huge for pet dogs. Think of all the independence you build as your dog solves problems in your presence with no need to seek help from you. Also, think about your dog's daily life outside of nose work; how often his independence is denied ("don't sniff about while we're on a walk") and how often you are his solution to a problem by way of your dog responding to your commands and controls to produce a reward from you.

Watch as many dogs & handlers do nose work searches as possible to build your observation skills - you learn a little from handling your own dog in a group class. You learn more from watching a few other teams in your class do searches. You learn a lot watching 25-50 teams coming from all over a region do searches. Start dedicating some time to watching teams work. Volunteering at events is a great way to get your observation fix, and it supports the sport!
  
Practice being a neutral observer when handling your dog, and when you do choose to interact with your dog in the search, be more of a responsive handler - a neutral observer allows the working dog the freedom to work and practices movement in the search that neither traps the dog in one area, or draws the dog away from an area without good cause. A responsive handler makes choices based on the dog's behavior (good cause), to support the dog's independent searching. Great teamwork can play out in many different ways, but it's ultimately the result of a very observant handler who trusts and knows his dog.

Remember, the handler doesn't necessarily have to be interfering in the search in an unusual way, other factors can create a difficult challenge that makes the dog more susceptible to handler cues that the dog would otherwise ignore in favor of doing his job. If you can identify these factors, you can set up training scenarios for your dog to learn from and hopefully overcome the same challenges in future searches. Every search should be a learning opportunity.

Create training scenarios for your dog that challenge his expectations and promote problem-solving to find odor - it's simply not enough for a nose work dog to think odor is equal to food. As one of the studies above mentions, companion dogs lack the independent problem-solving skills of working dogs, in fact, companion dogs often look to humans for help in a way working dogs do not. It is this understanding of companion dogs that makes it most important to create a learning environment for the dogs that allows them to build their problem-solving skills. What is a problem for a dog? Confronting change in their environment (it was here and now it's not), challenging their expectations (it was in the corners and on the perimeter of this room the last 10 searches, now it's not), and overcoming environment (how do I get through this narrow space/how do I walk on this wobbly floor/how do I navigate this rubble pile/how do I work with that noise in the background?). Make the dog a problem-solving, hunting machine, and odor importance will be easy.



If handlers can be such a drag on the dog, why be involved in the game at all? Well, as inferior as a handler can be when compared to his dog, the human end of the leash is actually quite vital. In training, you help the dog shape his natural talents to the human-chosen task. In competition, you possess important facts that help the team succeed under the conditions of the search (what are the boundaries of the search area, what is the search time, where have we gone or not gone, this is a container search or this is a vehicle search, etc.). And, probably most importantly, your dog wants to work with you! Yes, you provide a reward for finding source odor, but you're also your dog's partner, and he's excited to make discoveries with you. 

So, when will you get to the point where you're not a cue-producing factory on the end of the leash spewing out signals to your dog that challenge his ability to focus on finding source odor? The Dr. Seuss in me wants to say, "I don't know, go ask your pup." Most likely, it will happen in stages, over time. There will be progress and setbacks, highs and lows, and the learning will never end. If you find yourself doing searches with a deeper understanding of your dog's behavior, a deeper trust of your dog's independent efforts, and a less intellectual, more intuitive way of being in the search, you're definitely on the right track. Enjoy the nose work journey with your most excellent partner in sniff.   



Happy Sniffing!