Laying Down on the Job

Hi my dog Bella is very stubborn, she sometimes just sits down and refuses to do the obstacles even though I know she can do them. Any suggestions on how to stop/fix this problem?

Hmm… there could be a few different reasons!

Physically:

Is Bella really tired? If she’s been working a while, she may just be throwing her paws up with a ‘enough for today!’ If so, work her for less time and see if that fixes it.
Does Bella consistently lay down on certain obstacles? if so, what muscles do they work, or are they intense obstacles (such as a steep A-frame or high jump bar)? she could have a physical ache or pain keeping her from wanting to do an obstacle. Consider asking your vet to check her for tenderness, or do it yourself.

Mentally:

Are you 100% sure Bella knows how to do the obstacle? Try reintroducing it as a new obstacle and see what happens.
Is there something that could be scaring her about the obstacle? Did it bang? Did she hurt herself once? If so, recondition her carefully to accept the obstacle – maybe even rename it.

Is it rewarding?

Are you being rewarding enough? I don’t like to work for free, really. Volunteering makes me feel happy – that’s pay enough for me, but not for a dog. Dogs like their rewards to be tangible. Play, food, toys… try upping your reward schedule, and see if that helps.
If you’re rewarding enough, try changing your rewards. If you usually use food, try a toy.

Does anyone else have any ideas?

A lie down in middle of a training session can indicate a few things (hard to narrow the problem down without knowing the method of training…traditional compulsion training or clicker training method?). In general, some possibilities are:

– Session has gone on too long (dog is tired, needs to go to bathroom, is thirsty, etc.)
Solution: end on a positive note & try again after a rest, bathroom or water break, etc.

– Animal has injury
Solution: check it out

– It’s environmental (too hot, humid, something in the environment is making them uncomfortable such as other dogs, animals or people, etc.)
Solution: practice without these discomforts

– Session is not rewarding enough
Solution: increase value of the reinforcer (reward) and/or the rate of reinforcement

– Handler’s cue timing is off (a common problem in agility)
Solution: in order for one cue to mark/reinforce the previous cue in a behavior chain, such as that in an agility course, the cue has to be precisely given when the animal performs the preceding behavior in the chain or the chain breaks and animal goes unreinforced for all behaviors in the chain up to that point. For example, On bottom of A-frame, cue “tunnel”, just as their about to exit tunnel, cue “over” for the jump, as they’re going over the jump cue “walk on” for the dog walk, etc.

– Something negative is associated with that cue or obstacle: were they corrected or reprimanded? Did something scare them?
Solution: Re-shape the obstacle using positive reinforcement method such as clicker training and attach a new cue. The old cue might be poisoned.

– An extinction burst due to a faulty cue: dog doesn’t actually understand the cue given and might have erred by performing something else instead (and either didn’t get rewarded and/or were corrected). A faulty cue could also occur because handler isn’t doing what the dog expects (i.e., think of how they were trained, have they been shaped to perform that obstacle in presence of distractions such as variation in handler position/motion, and distance of handler, two aspects of a fluent behavior)
Solution: try a simple behavior for a quick reward, end on a positive note and rethink the training plan for that obstacle. Maybe the dog knows how to perform the behavior, but doesn’t understand the cue? Briefly re-shape behavior and re-introduce the cue. Strengthen the cue by generalization – i.e., vary their approach to obstacle from different directions, cue it with you in different positions/postures/executing different handler movements. If it is the behavior the dog doesn’t understand, break down the behavior into smaller achievable parts, re-shape with high rate of reinforcement and re-introduce the cue.

– The dog understands the obstacle (behavior) and knows the name of it (cue) but stimulus control isn’t quite there, That is, they do not perform the behavior every time the cue is given (one of the four rules that define stimulus control for an on cue behavior)
Solution: Investigate by cueing that obstacle only and randomly alternate between giving that cue and another well-understood cue (such as sit or a nose touch target of your hand…something that has a deep reinforcement history). Ignore mistakes but mark (say “good” or click a clicker as the behavior occurs if you are familiar with clicker method) and reward the correct responses for both cued behaviors. If they make a mistake, ignore it, & cue the easier behavior for an easy mark/reward to bring back their confidence/enthusiasm and re-cue the obstacle for another go. Track the number of correct responses, and missed cues. If their accuracy for the cue to that obstacle is 80% or above, their stimulus control is likely not the problem.

Note: If you have been using positive reinforcement in combination with reprimand or correction (or even handler frustration, which dogs do notice) this can frustrate/discourage the dog and cause what appear to be “refusals”. In that case, I recommend investigating the clicker training method.

Have fun with your Bella!

1 Comment on “Laying Down on the Job”

Pam Bromberg says:

July 25, 2013 at 3:03 am

A lie down in middle of a training session can indicate a few things (hard to narrow the problem down without knowing the method of training…traditional compulsion training or clicker training method?). In general, some possibilities are:

– Session has gone on too long (dog is tired, needs to go to bathroom, is thirsty, etc.)
Solution: end on a positive note & try again after a rest, bathroom or water break, etc.

– Animal has injury
Solution: check it out

– It’s environmental (too hot, humid, something in the environment is making them uncomfortable such as other dogs, animals or people, etc.)
Solution: practice without these discomforts

– Session is not rewarding enough
Solution: increase value of the reinforcer (reward) and/or the rate of reinforcement

– Handler’s cue timing is off (a common problem in agility)
Solution: in order for one cue to mark/reinforce the previous cue in a behavior chain, such as that in an agility course, the cue has to be precisely given when the animal performs the preceding behavior in the chain or the chain breaks and animal goes unreinforced for all behaviors in the chain up to that point. For example, On bottom of A-frame, cue “tunnel”, just as their about to exit tunnel, cue “over” for the jump, as they’re going over the jump cue “walk on” for the dog walk, etc.

– Something negative is associated with that cue or obstacle: were they corrected or reprimanded? Did something scare them?
Solution: Re-shape the obstacle using positive reinforcement method such as clicker training and attach a new cue. The old cue might be poisoned.

– An extinction burst due to a faulty cue: dog doesn’t actually understand the cue given and might have erred by performing something else instead (and either didn’t get rewarded and/or were corrected). A faulty cue could also occur because handler isn’t doing what the dog expects (i.e., think of how they were trained, have they been shaped to perform that obstacle in presence of distractions such as variation in handler position/motion, and distance of handler, two aspects of a fluent behavior)
Solution: try a simple behavior for a quick reward, end on a positive note and rethink the training plan for that obstacle. Maybe the dog knows how to perform the behavior, but doesn’t understand the cue? Briefly re-shape behavior and re-introduce the cue. Strengthen the cue by generalization – i.e., vary their approach to obstacle from different directions, cue it with you in different positions/postures/executing different handler movements. If it is the behavior the dog doesn’t understand, break down the behavior into smaller achievable parts, re-shape with high rate of reinforcement and re-introduce the cue.

– The dog understands the obstacle (behavior) and knows the name of it (cue) but stimulus control isn’t quite there, That is, they do not perform the behavior every time the cue is given (one of the four rules that define stimulus control for an on cue behavior)
Solution: Investigate by cueing that obstacle only and randomly alternate between giving that cue and another well-understood cue (such as sit or a nose touch target of your hand…something that has a deep reinforcement history). Ignore mistakes but mark (say “good” or click a clicker as the behavior occurs if you are familiar with clicker method) and reward the correct responses for both cued behaviors. If they make a mistake, ignore it, & cue the easier behavior for an easy mark/reward to bring back their confidence/enthusiasm and re-cue the obstacle for another go. Track the number of correct responses, and missed cues. If their accuracy for the cue to that obstacle is 80% or above, their stimulus control is likely not the problem.

Note: If you have been using positive reinforcement in combination with reprimand or correction (or even handler frustration, which dogs do notice) this can frustrate/discourage the dog and cause what appear to be “refusals”. In that case, I recommend investigating the clicker training method.

Have fun with your Bella!