How Many Repetitions To Teach A Required Behaviour?

[Staranais]

I'm very jealous of all you people who have dogs that take 5 - 10 reps to display understanding of a new exercise or concept! After years of patient training it still takes my own boy simply ages, probably dozens of short sessions, to display basic understanding of even the simplest new skill. (And no, it's not just me - I'm not the best trainer in the world by a long shot, but I'm not that bad. Even my cats picked up nose targetting way faster than my dog did!)

Obviously, like with humans, some of a dog's speed of learning new things can be accounted for by the dog's own innate intelligence. But I also wonder if lack of stimulation when a dog is young can affect how quick a dog is at learning when it is mature?

I can't help but wonder, if my boy had been socialised at all as a pup, if he had maybe even been taught a few simple commands or behaviours before I got him at 5 years old (even something as simple as house training), then perhaps he'd be "smarter" at learning now. Do you folks think there is maybe something akin to a critical period for "learning to learn" as a pup, like the critical socialisation period, and if no stimulation is recieved during this period then the dog will find it harder to learn as an adult?

Just thinking aloud, really.

Edited March 13, 2007 by Amhailte

[dogdude]

The brain is too complex to generalise. Look at the vast different learning capacity of humans!

Breed differences have also got to be considered. Bull breeds were not bred for there capacity to work closely with there human owners (like working breeds), so are not all that interested in changing their blueprinting when it suits us (obedience training). They are naturally going to take longer reps to get there, and you will have to find what really makes them tick to do it.

[Staranais]

Yup, believe me, I know.

But I've helped teach a few things to other bull breeds and other terriers belonging to friends/at the pound, and believe me, this guy is particularly slow even for a bull breed. He doesn't lack focus or motivation, he's just very slow to learn, even when he's focused and trying hard.

My dog wasn't socialised at all when young, wasn't taught a single thing until I got him at 5 years old, so I can't help thinking that these two things (lack of early stimulation and "slowness" at learning) are related?

Like I said though, I have no proof at all that these things are connected, so I could be dead wrong. Just thought other people might have read about or experienced something similar. ;)

[MonElite]

I think that if the dog has learned several words it is easier for it to learn the new ones. Basically the dog knwos the drill so to speak. You et shown what to do than you get rewarded. Thats why I think you can teach the old dog new tricks

[m-j]

Personally I don't have a set "rule" on how close the reps are. Much depends on the dog in question and its emotions at the time as to whether it needs tiny, small. medium or large breaks inbetween. Don't know what anyone else thinks on this.

Hi Erny ;) Definately you need to think about what you are working with imho if you don't you are making a rod for your own back but with your average dog if you only do one trial a day it would take even the smartest of dogs and trainers longer than 3-5 days to learn to sit imho

Sometimes this is too hard for novice dog handlers to do though,

;) not only novices *whistle whistle*

Even my cats picked up nose targetting way faster than my dog did!)

Don't despair dogs aren't actually that fond of doing proper nose targets, they quite often will place their muzzle (close to the nose) on the target especially when going for duration. This was told to me by a very experienced trainer of many different types of animals and is what I have found. Other animals I have done nose targets with I found much easier.

Did you find that you needed to show your cat where the treat was actually coming from ie between your fingers as they don't seem to be able to pinpoint where the scent is coming from as well as a dog can. I found my successive cats were much quicker as I taught the cat where the food was coming from initially before conditioning the marker, so that they could recieve the reward quicker.

cheers

M-J

[Staranais]

but still interesting - perhaps we need our own thread about cats!

Anyway, I taught them to target the end of a chopstick with their nose. Charged the clicker, then introduced the chopstick by putting it in front of their nose. Most cats will take a sniff at something new you wave in their faces, and that's when I captured the behaviour with a click. Took a few sessions before they got the connection (that they were making the click happen), but then I was able to up the criteria - I moved the chopstick so they had to walk to it, only clicked them sniffing the very end of the stick, etc.

I was impressed how quickly they both learned when compared to my dog (but then again, I was also impressed at the speed of learning when I taught nose touch to my friend's pit-x, so it probably says more about my dog than about my cats). ;)

Sounds like you taught it differently? I can't quite picture what you describe above.

[m-j]

I believe this can play a huge part in a dogs ability to learn or not learn, he spent the first 5 yrs (that is probably nearly half of his life) teaching himself what he felt he needed to know, probably with very little structured human intervention, you would imagine it would have to impact on his way of learning to some degree . Plus if a dog is overstimulated by the motivator they can try too hard and not get it.

- perhaps we need our own thread about cats!

Sounds like you taught it differently? I can't quite picture what you describe above

No the same way just my lousy explanation ;) While charging the marker when giving the cat the reward I held it with the tip of my fingers and suffered bitten fingers or once they had bitten the finger a couple of times they would sniff all around my fingers trying to locate the treat because the scent of the liver was also on the lower part of my fingers from picking up the treat out of my bag. Once they realised the treat was actually coming from the tip of my fingers it was plain sailing. Maybe I just need bigger treats so they can see them:) Dogs don't seem to have the same problem.

cheers

M-J

[Working_Setters]

I had to check in before driving to training.

Hi LL,

After a long day at work, I have to tell you that you're one lucky gal, you make me so envious ;) ;) That I too could give up work and spend my days dog training. Well I guess there's only another 30 odd years until I retire

[Flames_Daddy]

I think that if the dog has learned several words it is easier for it to learn the new ones. Basically the dog knwos the drill so to speak. You et shown what to do than you get rewarded. Thats why I think you can teach the old dog new tricks

This advice is GOLD people.

If you want to see some decent tricks then positive rewards are definitely the way to go. If I was a doggie then I reckon that's what I'd want bigtime.

I am still puzzled as to why the phrase "Gitoudavit" or "Get out of it" is so popular in Australia. Reminds me of Yosimite Sam is Looney Tunez Cartoons saying "Go on... Git.. ya Vaaarmint!!" Hmmm... but then again again he is a Hessian with Aggression issues. <inserts gunshot sound effects here for comedic drama>

[MrsD]

Teach = the dog recognises a voice & or hand signal & offers the correct response without luring.

Retention = being able to take the dog back to the same place that you taught the command X amount of time later & the dog will once again offer the correct response on voice or hand signal without luring.

Proof = when the dog will offer the correct response in any number of circumstances & places.

I agree with Myszka & others who mention that once you have "taught a dog to learn" it is often easier for the dog to learn new things. It also applies to things being taught that are related or similar to something that have already been taught. I was very interested in CTD's video of Moses doing the under & over thing whilst she did pushups. One can only assume that it was easier for her to teach that sequence because she has previously taught Moses to jump over her back. Whereas a dog which had never been taught to jump over anything would most likely take many more repetitions to do the same thing. (hope that all made sense )

[Lablover]

I had to check in before driving to training.

Hi LL,

After a long day at work, I have to tell you that you're one lucky gal, you make me so envious That I too could give up work and spend my days dog training. Well I guess there's only another 30 odd years until I retire

I am not spoilt, ...........I am ruined.

Mind you, life was not a bed of roses when my grandmother and mother were sick and since have passed away.

Life is very precious.

[Tangwyn]

Indeed it is. Savour every moment...

[Lablover]

Tell me about it. If someone can explain why indirect pressure works it would be a load off my mind!

Hi, owner/mother of lovely Nelson. Bringer of cold pizza, which I enjoyed at a trial one mid morning.

Indirect pressure. Bejesus girl, ............the amount of time it took me to understand indirect pressure. Send me your Email, and I have a description, even I easily understood.

Edited March 14, 2007 by Lablover

[NaturallyWild]

Lablover, i have some info somewhere at home about this from when i was overseas. Will track it down when i have a chance.

It all really depends on what you define "learned".

Obviously getting a repetitive behavioural response can occur within a single trial (eg you learn not to touch a hot plate after the first occasion, and you would learn to open a chocolate bar after the first time seeing and tasting one half opened - all animals are capable of this and its necessary for survival of the fittest). The trick to how quick this repetitive behaviour occurs depends on the motivation, complexity and similar experiences of the situation. If you can learn to setup situations where there is high motivation, low complexity and some similar compnents that guide the response, the behaviour will become repetitive quicker.

The other type of learning i would think about is that which learning that results in an "automatic" response - this is where the term "proofing" comes in to most people minds. The info i was thinking about indicated that it takes a minimum of 2000 trials (or something like that) for the neural pathways to have been laid down specifically for this.

Learning also occurs much quicker and is more permanent in the brain when the animal chooses the response to make rather than being forced, or lured too much into make the response. There are bichemical reactions that occur much higher when free choice is allowed.

Will get back with specifics when i find it.

ps. send me a copy of the indirect pressure stuff as well always love a read.

[Sardog]

Hi NW,

I'm not sure if you are referring to OS (Tom) in relation to your post but we were speaking to him a few nights ago and we discussed this topic at length. Similar thoughts and strategy. Border Collie 5-10 reps, Lab 10-12 reps, Golden 15 reps.

This test was done worldwide and data from 1,000 professional dog "trainers"...

Of course it all depends on what you are teaching/proofing and difficulty of exercise.

We must catch up ...soooon please.

[OzPetRescuer]

It would depend on the value of the behaviour to the dog no?

say I wanted my dog to push a button - if every time he pushed the button a treat fell out - and he was food focused - he would learn the behaviour quicksmart. If he wasnt food focused...then it wouldnt be as quick

Dogs do things to get what they want - quickest way to teach a behaviour is to make the dog think that what you and they want are the same things!

However, I also think you need to generalise a behaviour and wean off treats etc before it is well and truly learned - I guess it is up to individual goal identification

[Lablover]

How dogs learn is so much fun!!!!!

Now......to throw a spanner in the works, how many reps. does it take for a dog to learn an unwanted behaviour? Recently I had the pleasure of a very experienced US visitor who had trained many retrieving champions. We had many a discussion on the above.

Poll:

(1) more than a wanted behaviour

(2) less than a wanted behaviour

(3) the same amount of reps.

Edited: Brother, I wish I checked my typing before submitting.

Edited March 16, 2007 by Lablover

[OzPetRescuer]

a behaviour is a behaviour - if it is reinforced it will be picked up and kept

it just seems that unwanted behaviours are picked up more quickly because they annoy us - like my latest arrival learned from my senior cit to bark and scratch at the door for attention

[NaturallyWild]

How dogs learn is so much fun!!!!!

Now......to throw a spanner in the works, how many reps. does it take for a dog to learn an unwanted behaviour? Recently I had the pleasure of a very experienced US visitor who had trained many retrieving champions. We had many a discussion on the above.

Any behaviour (wanted or unwanted) can occur in ony a single rep - its just the motivation for the associated end result that repeats that behaviour.

Unwanted behaviour tends to get our attention so we respond to it, and dogs are extremely social animals so it is (a) very rewarding and (b) they made the choice for the behaviour which resulted in attention - why wouldn't you do the bad things if your a dog that is stuck a back yard and only paid attention for 30min at the end of the day or when your "naughty"

A lot of people also make the wanted behaviour to complex which means there will be a lot of hit and miss without reward - taking longer for their wanted behaviour to occur in the dog - compare opening a fridge and bringing back a beer to just bringing back a beer that is in front of the dog - very different tasks, but could each be considered a single behaviour. Be a splitter, not a lumper.

SARDOG - no it wasn't Tom i was refering to (but we may have discussed this over a few scotches - hard to remember), it was actually Suzzane Clothier, plus a talk at the APDT conference un the USA that i got some of that info. - Yessss have to catch up again soon.

Genetics definitely plays a part in the speed and intensity of the physiology that happens when learning occurs, but it also dictates preferences for the motivating factors eg border collie - movement, vs lab - food.

OzPetR - if you remove the treats (rewards) before the behaviour is well and truly learned - what is going to motivate the dog to do the behaviour?

The trick is in how the treats are given (or not given) most of the time (are they jackpots, and can you create non-contingent rewards, etc).

[OzPetRescuer]

NW - would never phase treats out til I felt the behaviour was well and truly learned

Say I have a dog - I want to teach it to press that button...

It does it within 5 reps

I go away then come back - dog immediately does the behaviour

I consider that learned - however there would still be generalising and proofing to go through to make sure the behaviour is really learned

A human example might be driving

you learn to drive - then you can drive - does that mean you dont get more adept at driving as you have more experience?

I think I just phrased myself badly in the last post - I can see how you would read it like that

I meant I would not consider a behaviour fully trained til the dog was weaned off treats and was able to perform the behaviour on cue in most situations without need for reward