Sunday, February 13, 2011

Clicker and Reward Marking



If you’ve been following my posts, you might have noticed that I believe that positive reinforcement done right, combined with non-punitive, brainy leadership is the fasted way to a reliably well-behaved dog – regardless of dog or problem behaviors. I don’t compromise on that, and luckily don’t have to even when my clients want to pursue group training I don’t offer anymore, cause in my neck of the woods are several like-minded trainers whose philosophy latch hand-in-paw with mine - and if you want to know who they are, email me.
Most of them are clicker trainers, and the common believe is that we all are; that positive reinforcement is synonymous with clicker training. But that is not so. There are some who apply the method without the clicker. I am one of them.
On that note, one could assume that, because the clicker concept in itself is incompatible with punishments, that all clicker trainers apply the reward-based method exclusively, but that is also not so. Some are incorporating the trendy gadget more as a calculated PR move than authentically shifting away from compulsion and correction training – so something to be aware of.

In case you have no clue what I am talking about, let me explain the clicker briefly. It is a small sound making device that marks the exact moment when the dog performs a desirable behavior, which gives the person a few seconds to hand over the real reward, which, with clicker trainers, is often a food treat. In other words, the clicker is a communication tool, with the sound clarifying to the dog that she’ll be paid for the brilliant thing she just did.
Why is reward marking with a clicker important for the layowner? Or is it? Should he even pay attention to such a technical detail?
In my opinion, reward marking is über-handy when shaping a new fancy trick, capturing a funky body movement, and in the beginning stages of teaching a new task. It can really speed things up, because the dog is not ambiguous what the bulls-eye action is. That is especially true if a frisky dog switches from one behavior to another quickly, or if the human is a tad slow in handing over the paycheck. So, a clicker makes a lot of sense if the correct moment is important but fleeting, cause it helps the dog to understand what exactly it is you are after.
A clicker can also be helpful for an owner eager to change from the traditional choke and jerk training method to positive reinforcement. Habitually used to notice the mistakes a dog makes so he can correct them, a clicker forces him to shift his focus to what his pooch is doing correctly which, of course, is the idea behind positive reinforcement training: teaching the dog a repertoire of behaviors that please us she can use to elicit what pleases her.
But in day-to-day life, or whenever you want prolonged connection and behavior duration, the clicker can actually present a hurdle.
For example, what I regularly see is an owner sending the pooch to her mat with the intent that she chills for a bit. The dog typically complies and lies on the mat alright, but instead of relaxing, tensely expects to be clicked and rewarded and, because she’s learned that the sound ends the exercise, self-releases right after. Not every trainer teaches that the click signals the end of the exercise, many of my friends don’t, but others do, and when they do, position duration stays with a dog who anticipates instant gratification can become difficult. If the click doesn’t manifest when expected, the dog becomes frustrated and fidgety, barky or whiny, and if it does, it is the dog’s cue that she can leave the mat and do what she wants, which often is returning to the behavior the owner sent her to the mat for in the first place.
We have similar problems regarding coming when called and eye contact. The recalled dog typically returns to her person, but doesn’t hang close for long and splits as soon as she’s received the click‘n’cookie, and rather than prolonged eye contact and true connection, the owner gets an automated quick glimpse. Eye contact attention is such a natural behavior for dogs that can so easily be fostered, that it is a shame to teach it as a trick, an artificial exercise.
I am also not convinced that a clicker is particularly useful in behavior modification. Here is my thing: if my imaginary three-year-old child is hyped to watch her favorite TV show, and I tell her to do something else first, and she does, and then I click her, fetch cookies and milk and then turn on the TV, would that decrease her anxiety, or would she be annoyed with the delay? Following that thought, if my fearful of dogs pooch stays calm and attentive to me when she sees another dog, the sensible and effective reward is to increase the distance to the other dog immediately, not to click, treat and then walk away.
Because kids are humans, it is easy for us to get what they want; with dogs that can be a little harder to figure out, especially for the more inexperienced rookie owner. There is a risk that the beginner clickerer focuses too much on the device, and not enough on what the dog’s motivators are. Yet, taking the time and interest to find out what drives the dog pays off, cause if she associates her person as the facilitator, her bond and connection increases, and with it obedience. The best behaved dogs are the ones whose individual and species-specific needs are understood and met.

Dogs, by virtue of their species, are very attuned to us and expert readers of our body and verbal communication signals. Dog owners, since forever, inadvertently or deliberately convey with a “good dog”, or a nod, when they’re pleased with their pooch’s behavior. The problem that delays training success is that many people stop with the praise, believing that to be good enough a reward. But if you follow it with a for your dog desirable consequence, your verbal “attaboy” or “yes”, your smile and gooey way you look at her, becomes a natural reward marker.
Clicker proponents argue that consistency is lost when we use our self to tell our dog that exactly this or that is what we want to see again, and yes, that is possible. And yes again, consistency is crucial in scientific studies, and important in our relationship with dogs and really anything we do in life, but I doubt that when I tell my dog with emotion, words and gestures how happy I am with her performance, that I adversely impact her future behavior. In addition, non-clicker reward marking might not be inconsistent at all. We all have an intrinsic and subconscious, yet habitual way in which we express ourselves that our dogs are, or at least should be, able to interpret.
A clicker-conditioned dog doesn’t have to observe the owner. I want my dogs to observe me. Our dogs connect the dots between their behavior, our human-typical responses, and what happens next. If we involve ourselves in a positive way, they’ll be wanting more of us. It is that voluntary, deeper connection I want to foster, and not an artificial approval signal.

Before clicker aficionados jump me, if you are successful with it, of course continue. If you and your dog are a team, and everything is cool, don’t change. And if you are looking for trainer, don’t rule her out because she clicks.
But if you feel that you and your pooch are stuck somewhere, it is not because you’re not correcting enough, or because she is dumb, defiant or dominant, but perhaps because she is on autopilot like a Skinner rat, and you might wanna rethink some aspects how you relate with her.
And you owners who embrace positive reinforcement but not the clicker, don’t worry that you're doing things wrong. Think KISS and Keep It Simple, Stupid. Use your voice and body to tell her how brilliant she is, and how happy you are that you partnered up; but do tell, and reinforce the desired actions with something she really wants, and you’ll be applying the reward marking concept very effectively.

Maybe, by promoting the clicker as one of the wonderful tools available to us, instead of the only or best one, more people might embrace positive reinforcement as the sole way to raise, teach and live with their canine companion.




1 comment:

  1. Great Blog as usuall

    Happy Valentines Day.
    Your biggest fan!

    ReplyDelete