Can Dogs Have Learning Disabilities?

[poodlefan]

Corvus:

Forgive me if I sound irritated. Designing this study has pretty much consumed my life for the past few months and I have looked at it from so many angles and taken in so many possibilities, but in the end the best I can do is try to guess what is most likely, design it for that scenario, and then just try it and see what happens. That's what I'm doing.

Perhaps if you had offered the purpose of your study, we'd not have gone off track. I assumed your premise had something to do with how dogs learned.

I understand that this is a scientific study and not dog training. However I think its a fair enough comment that you can't conclude a dog has a learning disability because a standard method you're applying hasn't worked. You have already modified your standard method for this dog and that didn't work either. I offered some possibilities as to why. So did others. My suggestons were targetted towards helping the dog 'get it', not your study results.

Your premise that shelter dogs would be no different to random sample of dogs is potentially flawed IMO. The primary reason dogs are offloaded in pounds is because they have unwanted behaviours.. behaviours usually controlled by socialisaton and training. My guess is that a random sample of dogs would find a higher proportion of dogs trained to offer behaviours with the expectation of reward than your test subjects. That would probably change your findings.

Maybe the dog doesn't like bonox.. ?? From a TRAINING perspective, I'd also suggest varying the reward. Different foods could produce a different result. A dog with low food drive might also respond very differently to a non-food reward.

Maybe the dog's just plain stressed.

Edited January 14, 2011 by poodlefan

[pipsqueak]

I'll ask a question that I'll probably be shot down in flames for, by you. Over what period of time were the 20 reps done? If, as poodlefan suggested the dog might have been stressed, doesn't that mean that there is a build of up cortisol in the system which takes time to disperse - I've read the current thinking is up to 24 hours - and that this cortisol can impede learning. Again, with our sighthound cross, you can do a dozen reps, which no apparrent success or association, take a break for a day or two, come back to the skill, and she is leaps and bounds ahead of where you left off...

[Aidan3]

Corvus, do you have an inter-trial interval in the conditioning phase?

[corvus]

Corvus, do you have an inter-trial interval in the conditioning phase?

No. The ITI comes in once the dog is touching on cue and we introduce a different signal for the discrimination training. I originally had it for the second phase, which involves training a cue for touching, but I ended up cutting it down to just a 15s pause because Erik was walking away. If he walks away from a training exercise, I know I've got problems! Bringing it in for the third training phase is causing a bit of worry as well, though, as suddenly the reward rate drops and dogs are waiting around a bit between trials. So far those that have made it to phase 3 are coping with the frustration, but it makes me a bit nervous to see frustrated dogs, even if they get over it on their own.

[Aidan3]

Corvus, do you have an inter-trial interval in the conditioning phase?

No. The ITI comes in once the dog is touching on cue and we introduce a different signal for the discrimination training. I originally had it for the second phase, which involves training a cue for touching, but I ended up cutting it down to just a 15s pause because Erik was walking away. If he walks away from a training exercise, I know I've got problems! Bringing it in for the third training phase is causing a bit of worry as well, though, as suddenly the reward rate drops and dogs are waiting around a bit between trials. So far those that have made it to phase 3 are coping with the frustration, but it makes me a bit nervous to see frustrated dogs, even if they get over it on their own.

I was wondering whether the Boxer had either learned to wait, or had offered the response during an ITI and was unsure what to do because of it. Definitely not the case then.

One problem you might have with shelter dogs is having them hungry enough for auto-shaping.

[pipsqueak]

Corvus, do you have an inter-trial interval in the conditioning phase?

Could you explain what this is, please?

[corvus]

I'll ask a question that I'll probably be shot down in flames for, by you. Over what period of time were the 20 reps done? If, as poodlefan suggested the dog might have been stressed, doesn't that mean that there is a build of up cortisol in the system which takes time to disperse - I've read the current thinking is up to 24 hours - and that this cortisol can impede learning. Again, with our sighthound cross, you can do a dozen reps, which no apparrent success or association, take a break for a day or two, come back to the skill, and she is leaps and bounds ahead of where you left off...

We do 15 minute training sessions broken up into 3 lots of 3 minutes on and 2 minutes off. Up to 2 sessions a day, an hour apart, and this dog did 4 sessions over 3 days, I think.

Impeded learning is a bit different to no learning. She may well have been stressed, but if she's still interested in a bit of puppy milk my guess is she's not so stressed her learning would be so significantly affected as to put her so far behind other dogs. She was a bit nervous about the marker tone to begin with, and she's the only dog we've had that has startled from the tone and then got over it. I wasn't going to say she had a learning disability, I just wondered if it was possible. Possibly I'm making a false distinction between her and some of the other dogs we've had problems with, but there's no way for me to tell, really. She makes me wonder and I wondered out loud in public to see what other folks may have experienced.

[poodlefan]

I'll ask a question that I'll probably be shot down in flames for, by you. Over what period of time were the 20 reps done? If, as poodlefan suggested the dog might have been stressed, doesn't that mean that there is a build of up cortisol in the system which takes time to disperse - I've read the current thinking is up to 24 hours - and that this cortisol can impede learning. Again, with our sighthound cross, you can do a dozen reps, which no apparrent success or association, take a break for a day or two, come back to the skill, and she is leaps and bounds ahead of where you left off...

We do 15 minute training sessions broken up into 3 lots of 3 minutes on and 2 minutes off. Up to 2 sessions a day, an hour apart, and this dog did 4 sessions over 3 days, I think.

Impeded learning is a bit different to no learning. She may well have been stressed, but if she's still interested in a bit of puppy milk my guess is she's not so stressed her learning would be so significantly affected as to put her so far behind other dogs. She was a bit nervous about the marker tone to begin with, and she's the only dog we've had that has startled from the tone and then got over it. I wasn't going to say she had a learning disability, I just wondered if it was possible. Possibly I'm making a false distinction between her and some of the other dogs we've had problems with, but there's no way for me to tell, really. She makes me wonder and I wondered out loud in public to see what other folks may have experienced.

Your other option Corvus is a dog that's been "disciplined" for nose touching.. food, plates, people.. Some folk aren't above back handing a sniffing dog. That would be those of the "get out of it ya mongrel" school of training.

I wonder if foot targetting would produce a different result.

Edited January 14, 2011 by poodlefan

[Aidan3]

Corvus, do you have an inter-trial interval in the conditioning phase?

Could you explain what this is, please?

Not very well When you have two interval schedules of reinforcement running concurrently to see how the animal makes choices, there needs to be a brief period of no reinforcement after each trial otherwise the animal will quickly figure out how to maximise reinforcement rather than making a choice that reflects the frequency or magnitude of the reinforcers on offer.

In plain english, it's a set period of time where nothing happens between trials that prevents cheating.

ETA: the first half of this article gives an overview of why we study choice behaviour and some of the things that influence choice

http://www.clickertraining.com/node/2970

Edited January 14, 2011 by Aidan2

[corvus]

No a random selection of pets wouldn't have been any better. A selection of dogs from the same kennel might have been a better group. However Aidan thinks it isn't a problem anyhow.

It depends on what you're interested in, though. If I were to take a sample from one kennel, all I'm learning about is that one kennel. If I were interested in something inherited, that would be a good idea, maybe. But if I want to know what the general level of optimism is in dogs, I should be aiming for a sample that is representative of the general dog population. I was originally thinking shelter dogs would not be representative of the general population, but I now think they are more representative than I was imagining. Most of the dogs are just typical dogs. The basket cases get PTS before we get our hands on them. Of the dogs left the worst they have displayed is timidity or a lack of training. I think if anything I'll end up with a bias towards optimistic dogs. Anyway, I'm still considering balancing my pilot study numbers with dogs from the public. We'll see what we end up with from the shelter.

Aidan, I think you're right, I think we do have some dogs not hungry enough. We already have one we are dropping because she works for about a minute, then just lies down. I've seen her at it and I am quite sure she would reach criteria if she ever worked for a whole training block, but she just doesn't want puppy milk that much. She's working for the hell of it, which really only sustains her for a couple of minutes max.

[sas]

I met this Boxer that after 20 reps of a very simple association still seems completely unaware of it. It seems unusual to me. Every other dog so far has picked it up in half that many reps or less. Is she stupid?? I've never said that about a dog before!

Depends what you class as a disability.

It would be completely logical to realise that all animal regardless of species learn in different ways and in their own time and that some may not be slower than others.

[corvus]

I wonder if foot targetting would produce a different result.

Good question. There was a time I considered going with foot targeting, but when I thought about it from an ethological perspective I thought it was more likely for a dog to explore an interesting scent or object with their nose rather than their foot. I'm always seeing dogs doing stupid things with their noses, like checking out an animal that is potentially dangerous. You'd think it would be safer to do that with a foot, but the only time I've seen dogs do it is if they are frustrated or trying to play with something. Anyway, given she's a Boxer she might be much more likely to use a foot than a nose.

Maybe that's what happened, but I'm more inclined to think if her behaviour can be explained by any past human interaction it was just people doing everything for her so she didn't have to think. She doesn't seem nervous at all. No sign of aversion or inhibition about nosing. We have a little dog that flinches every time a hand comes near him. That's pretty telling. This girl's a big cuddle muffin, though. Not in the least bit cautious around hands and more than happy to stick her nose in my lap/armpit/face/hands.

[Erny]

I'd love to video this. It's freaking fascinating. Never seen a dog look so much like they are getting it without actually getting it. The light is certainly on, but I don't think there's anyone at home. I imagine she would get there in the end, but I'm amazed and fascinated that she is so slow in comparison to the other dogs. Generally, they either pick it up in the first couple of sessions, or they don't interact with it at all.

The two dogs (sisters) I mentioned? One of them I basically all but gave up on the target exercise with her. It wasn't important for what I was doing and I only have the dogs for a short period of time with a lot to teach them, so I don't have a lot of time to waste on an exercise that is not part of what people necessarily want, even if it can be handy for the essence of learning. But as it turned out, I'd haphazardly give the exercise a go, quitting almost straight away if the dog showed no interest. Then there just happened that on one try (by this time me expecting it to be unsuccessful anyway) and the dog did it. After that, the dog would do a few repeats but I found I needed to keep the exercise really short with very few repetitions, but she did at least cotton on and did show enthusiasm at least for those first few reps.

So another thing I'd try (with all else apparently failing) is to quit the exercise rather than to keep trying in the one session and when you get a win, quit before you think the dog might. Similar to tug. Quit whilst the dog wants more. But I think that's the essence of training any way.

Edited January 14, 2011 by Erny

[poodlefan]

We have a little dog that flinches every time a hand comes near him. That's pretty telling.

Perhaps not. I sometimes think its got more to do with the relative size of the hand to the dog than any conditioned hand shyness.

Each of my poodles will duck their head to avoid a hand being placed on it - the Whippet doesn't care. None of them have been hit on the head.. unless you count the time I made Ted see stars by clobbering him with a kong on a rope.

Putting something the size of a garbage bin lid to a larger dog might produce a similar response in some I reckon.

Edited January 14, 2011 by poodlefan

[JulesP]

Why don't you ask in the boxer thread how boxers go with targeting in general. Especially with their noses.

[corvus]

Hey, incidentally, the star of the standard training methodology so far is a sighthound. He does the cutest 7 second touches, staring at the target and straining even his ears forward when he thinks the machine is broken and didn't register the first touch. He's gorgeous and is ripping through it with steady competence. He'd give Erik a run for his money.

[poodlefan]

Hey, incidentally, the star of the standard training methodology so far is a sighthound. He does the cutest 7 second touches, staring at the target and straining even his ears forward when he thinks the machine is broken and didn't register the first touch. He's gorgeous and is ripping through it with steady competence. He'd give Erik a run for his money.

Greyhound?

[Diva]

Despite their reputation as being difficult to train, sighthounds can be very good problem solvers. There was a youtube vid of some pranksters setting up dogs with a disappearing food bowl circulating a few months ago, not sure if anyone remembers it. Most dogs just look puzzled or wandered off but the only sighthound, an Italian greyhound I think, knew exactly what had happened and went straight to solve the problem. Very funny to watch.

[poodlefan]

Despite their reputation as being difficult to train, sighthounds can be very good problem solvers. There was a youtube vid of some pranksters setting up dogs with a disappearing food bowl circulating a few months ago, not sure if anyone remembers it. Most dogs just look puzzled or wandered off but the only sighthound, an Italian greyhound I think, knew exactly what had happened and went straight to solve the problem. Very funny to watch.

Yes, it pays not to confuse lack of interest in jumping through hoops for an owner with lack of intellect.. Deys good problem solvers those skinny dogs. Tick the WIFM box and watch them go at it.

[corvus]

Greyhound?

Indeedy. He has taken to it like a duck to water. I very much like watching him work. It's pure delight. He's a thinker. Yesterday he was caught on the wrong side of the apparatus when the tone sounded and he just leaned over the top of it to set it off from behind, then came 'round for his reward. Clever cookie! He doesn't just learn the task.