German Shepherds Forum banner

Studies suggest "less is more" when it comes to training?

2K views 19 replies 11 participants last post by  Deb 
#1 ·
Thought this was interesting... considering how most people (especially those involved in Schutzhund/IPO or in training LE and PP dogs) train, or believe they should train, their dogs in short sessions, multiple times a day, multiple days a week, while these experiments seem to show that dogs learn better when trained using short sessions, once a day, one to two times a week. So, we got the length that the training sessions should be down (short and sweet!), but it looks like we may have it wrong with how often we should be working our dogs? What are your guys' thoughts?

The first experiment involved 18 laboratory beagles divided into 2 groups.
Group A: Trained once a week
Group B: Trained 5 times a week
Each dog was trained to perform the same task using shaping and marker training and trained by the same trainer.
The results suggested that the dogs trained once a week (Group A) learned the task in fewer sessions than those trained 5 times a week (Group B). However, the dogs trained 5 times a week (Group B) did learn the task sooner than the dogs trained once a week, even though it took them more sessions.
Reference: Meyer, Iben; Ladewig, Jan (2008). The relationship between number of training sessions per week and learning in dogs. Applied Animal Behaviour Science, 111(3-4), 311-320.


The second experiment involved 44 laboratory beagles divided into 4 groups.
Group A: Trained 1-2 times a week, 1 session
Group B: Trained 1-2 times a week, 3 sessions in a row
Group C: Trained daily, 1 session
Group D: Trained daily, 3 sessions in a row
Each dog was trained to perform the same task using operant conditioning and trained by the same trainer. Each dog was trained for 18 sessions total.
The results suggested that the dogs trained 1-2 times a week (Groups A and B) performed better than the dogs trained daily, and that the dogs trained for 1 session at a time (Groups A and C) performed better than those trained for 3 sessions in a row. Four weeks later, all four groups retained what they had learned without much difference.
Reference: Demant, H., Ladewig, J., Balsby T.J.S. and Dabelsteen, T. (2011) The effect of frequency and duration of training sessions on acquisition and long-term memory in dogs. Applied Animal Behaviour Science, 133, 228-234.
 
See less See more
#2 · (Edited)
I'll admit, I didn't read the study (I plan too but it's late now). at a glance however... I'm not sure that this train of thought is all that new. think about folks who workout at a gym. everyday isn't a cardio day and when weight training, you vary the muscle groups.

if and when I'm training short multiple sessions daily, I'm not working on the same exercises.

if that's what's happening, no wonder pups get bored and pay more interest in whatever's over there instead of engaging with their handler.
 
#3 ·
Too many variables with studies like this. Yes they shaped behaviors and used a clicker, but that by itself isn't the whole picture when doing multiple sessions with a puppy. Then take into account breed differences what is actually being trained and if you're boring the puppy or not lol. I'll stick with what I'm doing.
 
#4 ·
Obviously these are very important factors that need to be taken into account, and I'd love to see more experiments conducted using different breeds, methods of training, and types of training (obedience, tracking, protection, agility, SAR, guide dog work, etc.), however I still thought that these experiments were interesting regardless.
 
#5 ·
Interesting study. I remember when my oldest son was young, they tried to set him up at age six to mentor some other children. He was very gifted and was used to 'teaching' his younger brother who was also very gifted and so learned the first time. He was exasperated with the children at school not understanding or learning as fast. I asked him who knew the material better at the end of two weeks, whether it was learned the fist time taught or the fourteenth time. He answered neither, they all knew it the same.


I think what I'm saying is whichever method a trainer is comfortable with or likes the best, it seems to work out in the end with the same results according to this study.
 
#6 ·
I usually train a pup once a week, during classes, while I am taking classes with them. They retain the material, they do good. Years later, they still have it down. I do think that some of us try too hard with our puppies.
 
  • Like
Reactions: gsdluvr
#7 · (Edited)
my job, in a sense, is my case study.


I am given 10 weeks to train 4 dogs (same age, breed, similar upbringing) the same behaviors / responses, including obedience and guidework. the dogs are worked (trained) 2x daily 5 days a week and they are tested at week 3 and week 8, then ultimately in class when they're paired with a visually impaired client & once that client returns to their home environment.

primary method of training is operant conditioning / clicker, but as each dog is different - their temperament and learning curve dictates your specific approach. the basic functions of a guide dog is to travel in a relatively straight line from point A to point B while avoiding obstacles and stopping at changes in elevation. in addition to these basic functions there are advanced lessons such as miscellaneous targeting (doors, chairs, hand rails, push buttons), traffic responses, sidewalkless techniques, etc.

the lessons are spread out as to not overwhelm the dogs, cause boredom, or progress too quickly - with the idea of retaining the information, reliability, generalization, and confidence. traffic training for instance is taught in stages, 3 sessions total, at week 4, 5 and 8.

each dog thinks and works a little differently - a couple samples of how each of my current 4 approach the same situation (curbs for example as the concept is pretty simple)

P - i get rewarded at curbs, let me find as many as possible

H - i stop at curbs otherwise Fodder trips and stumbles and that startles me

V - i stop at curbs because Fodder asked me to

also V - I'm still going straight, check! i moved to the left of that person, check! i didn't get distracted by that dog, check! i moved to the right of that dumpster, check! yay, the curb! finally the curb!

N - i stop at curbs because when i don't Fodder gives me a collar correction...i definitely prefer rewards.

because of these differences - i see it all over their learning. i did the same routes initially, patterning and clicking / rewarding each dog at the same points..... P was the most straight forward to train, she got it, there was an investment, some self motivation (ie fewer sessions). H received the same lessons, however the concept didn't sink in until i provided some consequences (same # of sessions, more repetition within the session). V had to develop a relationship, some trust, rapport, etc - for 3 weeks he had good routes then awful routes then good routes then awful routes then excellent routes and on and on. occasionally I'd skip a lesson with him as he has lower confidence, is slower to process but has the most willing nature (supports less is more theory) N needed a more traditional approach - corrections were introduced sooner - giving him the answer was not enough motivation for it to carry over and override his other interest and his natural drive to "just go" w/o thinking much. his sessions were the same but taught in a different manner.

in the end, as of yesterday, they all proved to have accurate and reliable curbwork as they passed their final guidework testing with a clumbsy blindfolded Fodder ;)

so, not sure if that interests you at all... but it's what I've got.
 
#9 ·
my job, in a sense, is my case study.


I am given 10 weeks to train 4 dogs (same age, breed, similar upbringing) the same behaviors / responses, including obedience and guidework. the dogs are worked (trained) 2x daily 5 days a week and they are tested at week 3 and week 8, then ultimately in class when they're paired with a visually impaired client & once that client returns to their home environment.

primary method of training is operant conditioning / clicker, but as each dog is different - their temperament and learning curve dictates your specific approach. the basic functions of a guide dog is to travel in a relatively straight line from point A to point B while avoiding obstacles and stopping at changes in elevation. in addition to these basic functions there are advanced lessons such as miscellaneous targeting (doors, chairs, hand rails, push buttons), traffic responses, sidewalkless techniques, etc.

the lessons are spread out as to not overwhelm the dogs, cause boredom, or progress too quickly - with the idea of retaining the information, reliability, generalization, and confidence. traffic training for instance is taught in stages, 3 sessions total, at week 4, 5 and 8.

each dog thinks and works a little differently - a couple samples of how each of my current 4 approach the same situation (curbs for example as the concept is pretty simple)

P - i get rewarded at curbs, let me find as many as possible

H - i stop at curbs otherwise Fodder trips and stumbles and that startles me

V - i stop at curbs because Fodder asked me to

also V - I'm still going straight, check! i moved to the left of that person, check! i didn't get distracted by that dog, check! i moved to the right of that dumpster, check! yay, the curb! finally the curb!

N - i stop at curbs because when i don't Fodder gives me a collar correction

because of these differences - i see it all over their learning. i did the same routes initially, patterning and clicking / rewarding each dog at the same points..... P was the most straight forward to train, she got it, there was an investment, some self motivation (ie fewer sessions). H received the same lessons, however the concept didn't sink in until i provided some consequences (same # of sessions, more repetition within the session). V had to develop a relationship, some trust, rapport, etc - for 3 weeks he had good routes then awful routes then good routes then awful routes then excellent routes and on and on. occasionally I'd skip a lesson with him as he has lower confidence, is slower to process but has the most willing nature (supports less is more theory) N needed a more traditional approach - corrections were introduced sooner - giving him the answer was not enough motivation for it to carry over and override his other interest and his natural drive to "just go" w/o thinking much. his sessions were the same but taught in a different manner.

in the end, as of yesterday, they all proved to have accurate and reliable curbwork as they passed their final guidework testing with a clumbsy blindfolded Fodder ;)

so, not sure if that interests you at all... but it's what I've got.
I think I fell in love with V. But then I was a Special Ed teacher until July when I retired. Two steps forward, one step back, lots of encouragement and Voila! *G*
 
#8 ·
In regards to a personal puppy. The number of times it takes a pup to learn a task is irrelevant for me. I would tell pup owners especially new puppy owners to train as much as possible. The pup is going to learn the tasks. What I find more important is the bond that me and all of my pups have formed while training. Sure we bond through play, feeding them, etc... But training sessions seem to be where the bond is formed the strongest. Or solidified so to say.
Now if I was training dogs for other people then this stuff would all play into my training schedule.
 
#10 · (Edited)
^ I'm in love with him too!!

eta: I also feel its my responsibility to clarify that the samples provided were simply to explain the differences in the learning process for the dogs and its a glimpse into a single aspect of their training. the approach is balanced and great attempts are taken to make sure that the dogs are enjoying and comfortable at their job. not operating out of fear as a "standard". afterall, happy dogs are how longevity in the work is created.
 
#11 ·
I think learning and a conditioned response, and body/muscle/emotional /mental conditioning are totally different... It is amazing how quickly kids learn how to wash dishes, but getting them to do it correctly every time, thoroughly even when they don't want to is different.. And they understand the "why's ' and' what for' that dogs don't...

Short repetitive sessions under progressive pressure and stimulation cause the learning to become seared into the whole animal and operant in even the most stressful of times.. I have been impressed at how quickly a sit and come command can be taught, but for the dog to do it every time especially when they don't want to or in an emotional environment is totally different...

Studies done in a laboratory is not real world... It is a controlled environment and subject to scrutiny imho
 
#12 ·
I think the learning is individual to the dog. I've found that when we are learning something knew, if I work it in short sessions for a couple of days and then just let it rest, Seger comes back stronger. What he learns never really goes away. It may have to be refreshed if we haven't done it in awhile but within minutes he has it back.

My girl is ADHD. We had to work on things daily. A couple days off and it was like we'd never done it before.
 
#13 · (Edited)
^ re: Seger, that's also why I LOVE testing on Mondays after the dogs have had the weekend off! a lot of ppl hate it but I'll trade a minor flaw here and there for a dog with enthusiasm and drive - it's almost as if they're excited to show you what they know. I also try to limit their exposure to testing routes so that it's novel and interesting to them.
 
#16 ·
I think you need to give them time to process. It's no different than us when we learn something. My friend puts the dogs up in a crate in a quiet place for an hour after training. That's based on some study that I'd have to ask her for. It gives them time to destress and think. That's similar to what I see in Seger when I train something new, give him the day off and restart. It's like they've had the "aha" moment while they are resting.

Do I think you only need to train one or two days a week? No. I think you'll be a long time getting there if you do that. But downtime is essential for them.
 
#18 ·
Yup. Learning and memory is a physical process.

I once had a dog come through training about a year or so ago from a rescue. It was this 8 or so year old hound named William. He was a little off but otherwise seemed normal except for one big area. I would train him how to do something and he would get really good at it that day but then the next day it was gone and we would have to start all over again from scratch. Funny thing was it wasn't completely from scratch the next day he would need to be shown it over again and he seemed to pick it up slightly faster than the day before, but he would legitimately forget and be unable to perform the behavior. Commands he knew from before he came into training stuck well, but anything new he didn't already know would be there one day gone the next. So I always just guided him. I made the decision I wouldn't hold him accountable because he was old and I felt bad for him thinking maybe he was just going senile. A few months later he went downhill super fast and died. It turns out he had a brain tumor and it was that brain tumor that stopped new information from going into long term memory. Sad, but I'm always glad I quickly decided to go really easy on him and let him coast through.

Every so often a dog will struggle with a concept only to have it perfect the next day after being able to sleep on it. Sleep plays a vital part in the physical process of learning and memory. Ultimately memories are a physical change in the brain on some level. If this physical change does not take place the brain does not retain the information.
 
#20 ·
People often get confused between teaching what are essentially tricks vs creating behaviours that are reliable and reproducible regardless of environment or distractors.

One is tricks and one is actual dog training. The reason most behaviourists cannot actually train a dog or fix behavioural issues.

Blitzkrieg1, I debated all morning whether to answer to this. But it shut down a very interesting discussion. I don't think you can paint all behaviorist with the same brush. Are they all equal? No. Just as with any method of training, some are better and some are not. I'll be honest, I'm a behaviorist in training. I changed over in the mid eighties when saying you were a behaviorist was the worst thing to say. I had a dog trained in obedience in the Koehler method. We didn't do well. I moved due to military move and the only person doing obedience introduced me to a new method. I retrained that dog using new words and we never went below 198 and even had two 200s. Later on I trained a collie in obedience who was never less than 196, who was a therapy dog and did herding. I was an obedience instructor for 25 years. I've had a SCH3 GSD. I've bred and showed dogs since 1992. I'm not a novice to dogs or training. Did behaviorist training work with every dog? No. Did it work with every owner? No. I've had Rottweilers and GSDs that I could take anywhere and they'd behave but their owners couldn't handle them. Would I have liked to place the dogs with people who could handle them? Yes. But that's not an option. I had to change methods and worked hard with the owners and the dogs to listen and they learned. It wasn't the best home for the dogs, but I couldn't do anything about that. I've had dogs go up the leash at me, a Mal I remember distinctly, and he got a very harsh correction. I'm sure I'm not the only one on here who uses behaviorism for training.


One thing I like about behavioristic training is the dog learns to behave for whoever has the lead, not just who trained them which sometimes is a problem with working dogs. I'm talking about family dogs. Any two year old could walk my collie or my past GSDs with the exception of Shana. I only trusted her with my own kids and my youngest did walk her.


What I'm saying is there are different methods of training. I don't think putting down a different method of training is conducive to others. I admire your knowledge and your ability to share it with others. You can read what someone writes and are able to get right to the heart of the problem.
Respectfully,
Deb
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top