Research into frequency of training

I think we all need time to process new information, new skills etc. whether human or dog. If I remember correctly, and approximately, what the study says is that dogs who have more frequent training sessions will acquire the behaviour quicker but the dogs on the once a week regime performed the behaviour better. I know both Juno and I get bored if we train the same things every day. Apart from early recall and sit/stay in the garden most of our 'training' has been an integral part of our daily walks, trips into town etc.

 

I try to do something everyday even if only for a short time. I also try to mix it up so I'm not always doing the same thing everyday . . . with one exception. Been working on the hold and drop command. Very tough and now on our second week. I have him sit on his "box" and do no more than 5 times then we're done. He's getting better at not using it as a chew toy but I think another week of work will be necessary. The hold and walk to heel is next step. The idea is short, sweet, and then off to something else. As I read one trainer state, "it's better to do 2 minutes of training a day than none at all."

 

@AlphaDog Yes, I would have said exactly that - better to do a short session every day - but the research paper I've linked to found the opposite!

Today I've discovered that there have been dozens of studies in humans on the same thing and with the same results i.e. people of all ages retain information better if not taught it on consecutive days. (So why did I spend many years as a teacher of young children insisting they should read every day?!)

 

It only found it with a very limited set of pretty much laboratory style conditions though, with a narrow new behaviour being trained, without distractions. And both the studies used Beagles! I train with some Beagles....jeez....they have the attention span of the average fruit fly....

 

Everyday, even walking or talk with friends in the street is a good time to train, specially for proofing and consolidating. We're training walking on leach carefully, and when we have distractions, i usually go to the "old stuff" (sit, up, come, down, etc).

 

We've been away without Molly for a week (back a few days ago) and I had such a good training session yesterday, working on off-lead heel, close turns, straight halts and 'obedience style' delivery of retrieves, in a park with distractions. She seemed really refreshed and keen to work after a week with plenty of exercise (my son looked after her very well) but no training.

Earlier in the summer we had a group visit our dog training club to do a taster session of flyball. Well we'd done half a dozen lesssons a year ago but stopped as Molly was so unenthusiastic. However, she clearly remembered and raced down the jumps, picked up the ball and raced straight back!

It's true the study was small scale, and it goes against my instincts, but I think for Molly less frequent sessions produce better results.

 

I listened to a podcast a few weeks ago about learning in higher education for humans. The speakers discussed how Advertising is really good at getting people to remember stuff. And they referred back to a study (I can't remember if it was a formal study or not) where adverts were shown at the Superbowl. What they found was this:

If an ad was shown during a break. Then another ad was shown. Then the original ad was shown again. That viewers did not remember the product when asked about it at a later date.

But if an ad was shown during a break. Then other ads were shown. The game continued. Then quite some time later the original ad was shown. That the viewers had a much greater chance of remembering the product at a later date.

They call this "spacing". And it's a really important aspect of learning in humans. The idea is that you need to recall from memory the thing you're learning in order for it to stick. If you learn it and regurgitate it, much like cramming for an exam and then answering all the questions, that you are less likely to remember it long term. But if you study a bit. Take some time out. Then come back to the topic and dig into your memory to learn again, you have a much greater chance of retaining the information long term. It's the act of recalling from one's memory that is so powerful in making things stick.

I don't know if this can be transferred to dogs, but it seems this is what the article was implying? (I've not read the article! I'm supposed to be working but taking a quick break by reading posts here!! Back to work now!)

 

I hope its OK putting my tuppence-worth in here, though its not strictly concerned with training frequency but with training a pup - even basic stuff - when he/she is an only dog.

Molly is our 1st experience of raising a pup on her own. We had hoped she would have the old cocker Alfie at least for her 1st year but sadly that hasn't happened. All our previous pups have arrived when there was already a mature and fairly well trained dog already in the household.

We have noticed a huge difference in training Molly, just simply getting her to understand what we mean. Previously pups followed the example of the older dog and picked up commands very quickly. It was very obvious this was going on and its only now, with Molly alone, that I realise how much work the other dog used to do and that we now have to learn how to train Molly and its requiring a lot more forethought than we ever had to apply with the others.

I always used to say, half jokingly that Max (springer aged 5 when BJ arrived) raised BJ. I now realise how true that was Right, time to get myself and Molly into training mode.

 

@MF I was interested in your comments about the ads and how humans remember. The original article is about spacing training in dogs, but is very small scale.

 

A while back I took an interest in memory and the techniques that could be used (got to the stage where I could memorise the order of cards in a shuffled pack in under 3 minutes). The analogy that is used in a lot of memory books that talk about the memorising process is that of spinning plates on poles (remember those old variety acts?) - you have to nip back and give the pole a bit of a shake now and then to stop the plate falling off but there is no advantage in going back too soon.

 

@UncleBob how long after memorising the order of cards did you provide the answer? Was it immediately or some weeks or even months later? The idea of "spacing" in learning is so that you will be able to recall the information a long time after. It's about accessing long-term memory rather than short-term memory.

But thinking of dogs, it seems we are after a Pavlovian response where there is no apparent processing of information; it's simply "cue" and "response". In fact, if I understand correctly, we are not wanting them to spend time processing. If I say "Come", my dog must respond by coming immediately. As for Snowie... I can see him weighing up the pros and cons. He responds immediately to me and makes a move towards me. Then he stops, goes the other way, comes to me, goes the other way (all this is happening in micro-movements, accompanied by a certain expression that he knows he should come to me but doesn't really want to). And then the call of the smell of a garbage bin filled with discarded food is just too appealing.

I do find the following interesting: Snowie has received presents from different people. Mostly stuffed toys. When a particular person comes to visit, he goes to get their present and parades around with it. Is that cue and response? It could be. Or it might be a different type of recalling from long-term memory. (I don't know enough about the mechanics of memory -- clearly I've forgotten much that I learned in Psych 101! In those days it was cram and regurgitate!)

 

Sorry posted the above unfinished.
I think you make an interesting point MF and in the end we want most (all?) behaviour to be an instant response. (Does some behaviour always require thought, accessing older memories? Like finding a particular object among others when directed?) While training some behaviours, like sit, doesn't seem to need thought, whereas others need the dog to think while learning. I'm trying to teach Molly to line her body up with mine in a close heel. She's looking at me and definitely concentrating but is as yet unsure what I want, so she has to try very slight adjustments to see what will give the reward.

 

And yet @Joy I read on a Guide Dog thread that guide dogs do indeed need to think about what's the best action to take. For example, can I cross the road now when my handler is saying "Forward" but there's a car coming?

I do love how we can see Snowie is deciding what he wants to do -- when people come to visit, he goes to his bed and you can see he's deciding which toy to take to them, often selecting a high-value toy for a high-value person (unless he's trying to match the present to the person, as I said in a post above). It gives him character, shows he is not robot. But, on the other hand, I would love a robotic response to Come! (And yes, I am well aware that the Pavlov's response is all about training training training!!! )

 

Yes, after I'd posted, guide dogs came to my mind too. It would be interesting to know about their training schedules once past the puppy stuff. @Boogie any info please?

 

There are many examples where the dogs make choices and training is about helping them learn to make good choices. Other things like recall or stop we want as conditioned responses that are basically automatic.

Then there's the question already raised of learning a new skill vs proofing that skill.

Many variables to take into account and so much to train we need never do the same thing two sessions in a row if we don't want to

 

The most successful training (with Charlie, anyway) is where I can engineer a fast, complete response and associate that immediately with a cue. Ping pong recall has this characteristic. The action of throwing the treats quickly gets a whip lash turn going, and then that fast, complete response can be paired with a cue right away.

This kind of thing is aways the most successful. Anything that is shaped, or takes time (e.g. the dog is thinking and processing) is always much harder to to turn into a super fast response I think.

I suppose the point I'm thinking about is the closer you get to exactly what you want as a conditioned response right away, and the less processing the dog does the better, I think. I think this is why luring is much better than shaping (for example).

 

Do you think that's why shaping is falling out of favour?
I just can't seem to lure Molly into exactly the correct heelwork position. I wish there were some training videos of people working with dogs who don't immediately 'get it'. Most videos have a trainer demonstrating the early stages with a dog who already knows the movement. (Going a bit off the subject here.)

 

Yes, Guide Dogs do need to think and decide all along. 'Do I manoeuvre round the obstacle or wait?' 'Do I go forward when told or refuse and stand firm?' 'Is there enough room for us both to the side and is there enough heard room?' 'Is there a free chair to indicate?' etc.

I'm not sure how all these things are trained, but I do know that they breed and select dogs for enthusiasm and an independent streak.

 

That describes all my labs.