Teaching the Sit-Stay

In the past, dogs I've trained have really struggled with two aspects of the SIT-STAY: (1) the handler's turning her back on the dog to walk away and (2) the finish, in which the handler walks counter-clockwise around the dog at the end of the exercise so she ends with the dog in HEEL position.

Reading and watching material by Patricia McConnell recently has encouraged me to think of alternate methods to encourage dogs to perform as I'd like. Since I'm not taking on any new clients due to my illness, I've been itching to put these ideas into action

Now that we have Winston, a one year old rescue Pug (now neutered) who didn't even know SIT when we welcomed him to the pack, it is the perfect time to test these new--at least new to me--concepts.

STAYS are naturally difficult for dogs; they want to be close to their handlers, who love them, play with them and provide for their every need. The idea of allowing the handler to walk away is completely foreign to them. In some dogs, it can even be anxiety-provoking.

Traditionally, elementary-level STAYS are taught with the handler backing away from the dog, maintaining eye contact and taking a step or two while giving a hand signal for STAY (usually the palm-up "STOP SIGN"), then rushing back to reward the dog before he stands up. As the dog begins to understand the exercise, the handler either takes more steps or waits a slightly longer period before returning to the dog with a reward.

In my experience, the behavior first falls apart when the handlers turns his back on the dog to take a step or two after giving the command and the hand signal. So, it is back to baby steps until the dog understands that STAY means the same thing whether he can see his handler's face or not.

Once that has been reestablished, the next place the exercise falls apart is when the handler attempts the traditional obedience finish. Most dogs will twist their heads to follow the handler's movements as she walks around the dog, standing when the handler is out of sight.

To me, this seems like a lot more un-teaching of bad habits than teaching of good ones. So, with Winston, this is our step-by-step protocol.

(1) He learned SIT consistently--meaning that he sits on command with either verbal command or hand signal in a minimum of 8/10 trials with no visible food reward.

(2) To encourage him to sit for longer periods, he received food rewards every few seconds as long as all four feet and his butt remained motionless on the floor. If he stood up, I used my negative reinforcer ("AH!AH!") and guided him back to his original position.

(3) Once he offered longer sits (5-10 seconds) in a minimum of 8/10 trials with a consistent food reward, I varied the frequency of his food rewards.

(4) When he was offering the 5-10 second sits in a minimum of 8/10 trials without food reward, I introduced the command "STAY" with variable reward rates.

(5) When Winston consistently offered 5-10 second SIT-STAYS without food rewards, I began to turn my back on him. If he didn't move, I turned around to offer food rewards each time.

(6) Once Winston was accustomed to holding his stay for 5-10 seconds while my back was turned to him at least 8/10 trials, I began to take a single step away from him. Then, I turned to face him, took a step toward him and rewarded if he held his stay.

(7) When he was consistently holding his stay at least 8/10 trials with a variable rate of reinforcement, I slowly lengthened the number of steps I took away from him. With each additional step, I waited until he was consistently holding his stay a minimum of 8/10 trials with a variable rate of reinforcement before adding another one.

(8) Once I could walk away from Winston approximately 5 feet, I began to work on the finish. As I approached him, I side-stepped to my right and held a treat in my left hand at Winston's nose-level, repeating the STAY command. While I walked behind him, the treat remained in the same position, encouraging Winston to hold his stay until I stopped on his righthand side.

Winston was a quick study, but some dogs may need step #8 broken down further. If your dog has a really hard time with allowing you to circle behind him, start with baby-steps:

     (a) When you return to your dog, stop directly in front of him.

     (b) Once he holds his stay in 8/10 trials with you standing directly in front of him, hold a treat in your left hand and take one large sliding step to your right and stop.

     (C) When he consistently holds his stay in 8/10 trials with you standing to the right, begin to take a step forward (toward the rear of your dog). It will probably help to hold a high value treat at nose-level for this.

     (d) Once he holds his stay in 8/10 trials with you standing beside him, slowly slide around behind him, using a treat at nose level to encourage him to stay put as you move.

     (e) When he consistently holds his stay in 8/10 trials with you standing behind him, slowly move into HEEL position beside your dog, using a treat at his nose level to make sure he doesn't break his stay.

Once your dog has the basics down as described above, it is fairly easy to exend the amount of time the dog will maintain his position and his distance from you as long as you don't move too quickly.

Remember, slow is fast when training a dog. Asking a dog who can STAY for 15 seconds to jump to staying 2 minutes is unrealistic and the behavior will fall apart. Work up to a 2 minute STAY by increasing the stays a few seconds at a time.

Also remember that dogs do not generalize well. One who performs flawlessly in your living room may act like he's never heard the command at the dog park or at your mother's house. So, practicing in a variety of environments is key.

Start with the lowest level of distractions possible and work your way up.  Don't compare your older dog with a new one. My 5 year old Service Dog, Mickey, will stay if asked, even when there are other dogs present or if a rabbit runs past him. At this point, if a fly enters Winston's peripheral vision, he's going to forget to STAY. 

Above all, please make training FUN for your dog. 3-5 minute sessions several times a day are more effective than one long session. If you end while your dog is performing well, he'll be happier to train during the next session.