Luring Vs 'guide, Show, Place'

[Aidan3]

No, by definition and by function it would be an extinction procedure and not negative punishment.

I thought you would say that .

So that I can be clear with what your word definitions are, would you mind giving me a simple example of what you would define as Negative Punishment?

Negative punishment is when you remove a stimulus contingent upon a response, and that response diminishes. By contrast, Extinction is where a previously reinforced response is no longer reinforced.

In -P you are taking away a stimulus, in extinction you are failing to provide a reinforcer for a previously reinforced response. And that's how we avoid putting too fine a point on it :D

I wouldn't argue that either was any less stressful, that would depend on a number of factors; but the natural process of free operant shaping is to reinforce something else and that immediately mitigates stress.

[haven]

OT, just out of curiosity, how come you have two logins Aidan? Sorry if I've missed where you have explained it before but I tend not to spend so much time in here as I used to.

[Erny]

Negative punishment is when you remove a stimulus contingent upon a response, and that response diminishes. By contrast, Extinction is where a previously reinforced response is no longer reinforced.

In -P you are taking away a stimulus, in extinction you are failing to provide a reinforcer for a previously reinforced response. And that\'s how we avoid putting too fine a point on it :D

I don\'t want to put a fine point on it either - we\'ve done that before in other threads, LOL. Are you able to provide a simple training example of negative punishment?

I wouldn\'t argue that either was any less stressful, that would depend on a number of factors; but the natural process of free operant shaping is to reinforce something else and that immediately mitigates stress.

I agree. And it would probably surprise many as to how that stress would and could compare with other training methods.

Edited May 24, 2010 by Erny

[Aidan3]

Are you able to provide a simple training example of negative punishment?

Sure, the most common example would be pup playing with owner, pup gets too rough, owner walks off and shuts door behind them. The stimulus removed is the owner. Another example, dog starts barking while playing fetch, owner takes ball and puts it away.

I wouldn\'t argue that either was any less stressful, that would depend on a number of factors; but the natural process of free operant shaping is to reinforce something else and that immediately mitigates stress.

I agree. And it would probably surprise many as to how that stress would and could compare with other training methods.

I'm not aware of any empirical data that really looks at this so it would be hard to make any sort of judgement, though many do. My philosophy is that the dog should feel in control of the consequences. We do have empirical evidence that this is important. Arbitrary judgements about which "quadrant" a trainer will use I find a bit pointless in this regard.

[corvus]

I think there's a lot more to selecting a training approach with a particular dog than just how stressful the approach is likely to be. More important, I think, is how the dog is able to cope with that stress. I can put Erik through stress in training and he'll work harder, or he'll slow himself down to get more control of his movements. I have to be more careful with Kivi. Also, you have to remember that just getting out of bed in the morning sets off a very small stress response. Here's the abstract from an interesting paper showing how stress can affect learning through shaping or classical conditioning:

Rapid shaping of behaviour associated with high urinary cortisol in domestic dogs

Emily-Jayne Blackwell, Alina Bodnariu, Jane Tyson, John William Stephen Bradshaw, Rachel Alison Casey

a b s t r a c t

The occurrence of stress has widely been associated with impairments in learning abilities in animals, although the influence of stress appears to differ with the complexity of tasks. Previous research has suggested that some domestic dogs exhibit both physiological (elevated cortisol) and behavioural signs of stress when newly admitted to re-homing centres. In this study we have investigated whether levels of stress as measured by urinary cortisol: creatinine is sufficient to impair the learning of simple associations. On the day following their admission to a re-homing centre, 32 dogs were trained on one classical conditioning task and one operant conditioning task; 6 days later, they were trained on a second operant conditioning task. Their mean urinary cortisol:creatinine ratio (C/C) fell from 27.1 × 10−6 to 22.3 × 10−6 (nmol/l:nmol/l) between these 2 days; a substantial proportion (78% on day 1, 63% on day 7) of dogs had ratios above the range of 5–20 × 10−6, which is that defined as clinically normal (Bush, 1991), suggesting high levels of stress. The dogs’ average time to reach criterion on either task on day 1 was unrelated to C/C or to behavioural signs of stress; this lack of correspondence may reflect the diverse previous experiences of the dogs. On day 7, the six dogs which failed to reach criterion for the operant association were significantly less active and interactive in their kennels than the others. For the remainder, a high rate of learning the operant association was associated with high C/C (in excess of 40× 10−6), and a poor performance was associated with fearful behaviour in the kennel. Dogs appear to have adopted one of two coping strategies: either the display of fearful behaviours and an impaired ability to learn the tasks, which may reflect a ‘reactive’ style of responding, or a higher level of HPA axis activation and an enhanced ability to rapidly learn a new task, which may be indicative of a more ‘proactive’ coping style.

Edited May 24, 2010 by corvus

[m-j]

So, what do you tend to use when training dogs?

Any method/technique that is positive, mainly luring and targeting.

Why do you find that technique is preferable?

I like the idea of rewarding the only thing they can do right as opposed to making them at best feel uncomfortable about not understanding what I want them to do. I know not receiveing a reward can be stressful to a dog but that I use that as a learning curve for me and just go for a smaller increment.

Would you ALWAYS use that technique no matter what or would you move to the other if you think it would work better for the dog?

No I will change positive techiques. In the past I have gone back to the GSP technique but not for a long time I generally end the session and go and rack my brain as to how I can make the dog work better. Even with dogs that won't eat, my thoughts are that all dogs eat I just need to think about why they won't eat in a training situation or why they won't take a "freebie" treat off me etc. I know this can take a bit longer but to me training is not about how fast the dog can achieve something, but how the dog perceives the training and what long term results will be like. Plus I like a challange at least if I'm determined to do something using positives only I won't be stressing the dog out too much, being exposed to small amounts of stress is a good thing, but more importantly I learn from it.

I am one of those anal retentive positive trainers (I'm a crossover trainer and you know what reformed whatevers are like) but I certainly don't think traditional trainers are cruel (if I did I wouldn't have many friends left as in this area it is still very popular, although change is happening ) only when it goes to extremes and the dog receives a huge correction for not understanding.

But I like to introduce a combo and then GSP only. Which means the dog will learn each, singularly and a combo. Shaping I will use if necessary.

Teaches the dog to follow the hand and also teaches the dog it's ok to be 'put'.

I agree with the highlighted but I still prefer to teach that as an individual exercise not when I'm teaching a change of topography of the dog with a cue.

Also, you have to remember that just getting out of bed in the morning sets off a very small stress response.

Hmm I can't agree with that I don't like leaving my nice warm bed on these cold mornings

cheers

M-J