Kay LaurenceKay Laurence (Learning About Dogs) · 글
targetsTossing practise(던지기 연습), pointing practise(지시 연습), placing practise(배치 연습).Start / Stop cues(시작/멈춤 큐)를 알고 있는지.take from hand(손에서 받기)와 follow hand(손 따라가기, luring 시)의 차이를 알고 있는가?rewards(보상)를 얻기 위함임.Schultz Antecedents (슐츠의 선행 요건) 및 학습 공간resting areas(휴식 공간)와 learning areas(학습 공간)를 분리해야 하는가?Cue (큐) 기술cue를 계획했는가?cue를 전달하는 방법을 연습했는가?pre-cues(사전 큐)를 인지하고 있는가?cue를 제공할 때(즉, 개가 cue seeking을 할 때) cue를 전달하는가?cue를 반복하지 않을 만큼 절제하고 있는가?Transferring the cue (큐 전이)cue(예: target이나 lure)를 인지하고 있는가?cue로 옮겨가는 절차를 이해하고 있는가?performance cue(수행 큐)를 계획했는가?cue를 추가할 수 있는가?History (학습 이력)antecedents(선행 요건)와 관련하여 과거에 어떤 연관성(associations)을 맺고 있는가?The Environment (환경)Pre-cue (사전 큐)pre-cues가 존재함.cue가 시작되는 것보다는 pre-cues를 원함.pre-cues는 환경으로부터 발생할 수 있음.The Relevant Cue (관련 큐)Emotional context (정서적 맥락)cue(open session cues)를 가르쳤는가?cue(temporary end the learning)를 가르쳤는가?cue가 있으며, 이것이 개에게 실망감을 주지 않는가?Recognise (인지)cue(들)를 인지할 수 있는가?cue를 명확하게 볼 수 있는 위치에 있는가?cue를 보았다면, 그것이 어떤 행동과 짝을 이루는지 기억할 수 있다고 확신하는가?Cue seeking (큐 탐색)cue를 찾으려 하는가?cue seeking을 가르치고 보상했는가?Antecedent presented Dog Responds (선행 요건 제시와 개의 반응)Pace (반응 속도): 반응의 속도는 기억의 용이성, 반복의 욕구, 무엇을 반복해야 하는지에 대한 이해를 나타내는 결정적인 정보임.fluency(유창성) 또는 latency(잠재기)라고 부름.dog (반려견)에게 있음.mental fatigue (정신적 피로)를 나타내는 지표임.suppression of behaviour (행동 억압)를 피하는 방향인가?absence of behaviour (행동의 부재)를 추구하지 말 것.learner (학습자)에게 윤리적으로 적절한 방식으로 교육할 것.Guidance (지도)를 얼마나 제공할 것인가에 대한 기준을 정립해야 함.lure (유도물)를 따라오도록 가르치기 위해 손이나 cup on a stick (막대기에 달린 컵)을 사용하는 것에 능숙한가?following (따라오기)을 유지하고 biting (물기)이나 loss of focus (집중력 저하)를 방지하기 위해 lure의 속도와 engagement (참여도)를 조절할 수 있는가?lure/follow hand (유도/따라오는 손)를 take-it-now hand (보상을 즉시 주는 손)로 전환하여 mark (표시/클리커 등을 이용한 신호)를 할 수 있는가?self-directed (자기 주도적)인 상태가 되는 것에 익숙하고 자신감이 있는가?extinction (소거)을 피할 수 있는가?repetition series (반복 시리즈)를 구축하여 strength (행동의 강도)를 높일 수 있는가?touch-prompting (접촉을 통한 촉진)을 사용하거나 환경을 이용해 반응을 제한하여 행동을 moulding (성형)함.voluntarily sought (자발적으로 탐색)하게 하는 것과, pressure (압력)가 implicit (암시적)으로 작용하는 것의 차이를 평가해야 함.Target acquisition (타겟 습득): 행동, 연상, 또는 동작이 어떻게 학습되는가?targets (타겟)을 prompts (촉진제)로 활용함.Replacing the target (타겟 교체) 과정이 필요함.Connection (연결)이 우리의 base behaviour (기초 행동)임: 학습이 시작되는 지점임.Connection은 반려견의 needs (필요), limitations (한계), abilities (능력)를 인지해야 함.Connection은 피로가 시작되는 시점을 알아차릴 것을 요구함.consistent starting position (일관된 시작 자세): sit, down, stand 등.consistent location (일관된 위치): 보호자와의 상대적 위치(정면, 측면, 거리 등).natural state (자연스러운 상태)에서의 행동 기능은 무엇인가?reinforcer (강화물)는 무엇일 가능성이 높은가?component (구성 요소)인가?component behaviours (구성 행동들)가 있는가?reward (보상)가 이 행동을 보완할 것인가?flexible (유연)한가? (단, variable 변덕스러운 것과는 다름)pool of the whole behaviour (전체 행동 덩어리)를 나중에 정리하려 하지 말고, 완성된 행동의 small slice (작은 조각)를 목표로 할 것.veneer (겉치레)를 덧씌우는 방식이 아니라, 반려견 내부에서부터 발달하도록 허용해야 함.rewards (보상)를 바탕으로 구축된 행동은 자연스럽게 strength (강도)를 얻게 됨.cue (신호) given when the dog was cue seeking (신호 탐색 중) / ready to respond (반응할 준비가 되었을 때)였는가?cue (신호) clear, consistent and unambiguous (명확하고 일관되며 모호하지 않았는가)?default (디폴트 행동) when uncertain (반려견이 불확실할 때 자신 있게 default 행동을 제시하는가)?Default (디폴트 행동)default (디폴트) behaviour which the dog should perform if they do not recognise the cue (신호), or would prefer not to respond? (반려견이 cue를 인식하지 못하거나 반응하기를 원치 않을 때 수행해야 할 default 행동을 가르쳤는가?)cue (신호), respond with the associated behaviour should not be questioned. (cue를 인식하고 관련 행동으로 반응하려는 반려견의 노력은 의심받아서는 안 된다.)learning history에 충실할 뿐이다.)learning gap (학습 격차) should be examined. (반려견이 기대하는 품질이나 수준으로 반응하지 않는다면, learning gap을 검토해야 한다.)cue (마커 신호)marker (마커) as a cue (신호) is learned by the dog. (cue로서의 마커는 반려견이 학습하는 것이다.)marker cue (마커 신호)s: (마커 신호에 대한 반응을 명확히 해야 한다:)cue (신호) to take it (보상을 기다릴 것인가, 장난감을 낚아챌 것인가, 아니면 가져가라는 cue를 기다릴 것인가)Targets (타겟)Target acquisition (타겟 습득): how is the behaviour, association or action learned (행동, 연합, 또는 동작이 어떻게 학습되는가)targets (타겟) as prompts to teach new behaviours (새로운 행동을 가르치기 위한 프롬프트로 targets 사용하기)target (타겟 교체하기)marker (마커) produce the accuracy in the repetition (선택한 마커가 반복 과정에서 정확성을 만들어내는가?)marker (마커) suitable for the behaviour - does the dog need to see the marker (마커) or hear the marker (마커). (마커가 해당 행동에 적합한가? 반려견이 마커를 봐야 하는가, 아니면 들어야 하는가?)Clicker (Clicker를 사용하는 이유)Markers (마커) are cues (Markers는 cues이다)marker (마커) should be paired with reinforcement processes (보상 과정과 연결되어야 한다).markers (마커). (특정 마커에 따라 다른 보상 과정을 사용하도록 선택할 수 있다.)marker (마커) is a cue (신호) to orientate and prepare to begin the reinforcement process (보상 과정). (마커는 reinforcement process를 시작하기 위해 방향을 잡고 준비하라는 신호이다.)Marker response (마커 반응)marker (마커) and evidence that they desire our reward? (반려견이 마커에 반응을 보이며, 우리가 주는 보상을 원한다는 증거를 보이는가?)reinforcement (보상) is likely gained from the environment. (만약 반려견이 반응은 하지만 보상을 찾지 않는다면, 환경으로부터 reinforcement를 얻고 있을 가능성이 높다.)marker(마커)에 대한 반응이 없으면 reward(보상)를 전달하지 않아야 함.Delivery cue(전달 신호): reward 과정에 대해 개에게 정보를 주는 명확한 cue(신호)나 패턴이 존재하는가?food(음식)를 '침대에서 아침 식사(breakfast in bed)'처럼 줄 것인가, '받아먹게 던져(thrown a catch)' 줄 것인가, 아니면 '바닥에 놓아(placed)' 줄 것인가?cue에 반응하는가?Variety of markers(다양한 마커): 다양한 marker를 사용하는 경우, 개가 각 marker에 올바르게 반응한다는 증거가 있는가?marker가 서로 다른 reward process(보상 과정)를 의미하는가?No cheer leading(응원 금지): marker(결과물의 시작) 이전에 cue 외에 어떤 것이든 제공하는 것은 해당 behaviour(행동)의 약점을 숨길 수 있음.cue로 요구하게 만든다.)marker가 발생한 후에는 얼마든지 큰 소리와 진심을 담아 응원해도 됨.Rate of reinforcement(강화율)는 다음 중 하나와 관련이 있어야 함.behaviour cycles(Antecedent, Behaviour, Consequence / 선행 조건, 행동, 결과)의 수.behaviour의 수.Rate는 behaviour 간 혹은 학습자 간의 비교가 아니라, 반드시 각 behaviour 별로 측정해야 함.reinforcement process가 완료되거나 consumption(섭취)이 일어난 후, 개가 다시 cue를 찾으려고 시도하는가?cue를 찾는 시간이 지연된다면, 이는 mental fatigue(정신적 피로)의 징후일 수 있음.Continuation(지속)은 동일한 behaviour를 반복하는 과정임.cue를 추가하여 유지(maintenance)를 돕고, 이후 뒤에서부터 앞으로(from the rear, forwards) cue를 점진적으로 제거(fade)할 수 있음.cue1, cue2, cue3, cue4 → cue1, cue2, cue3 → cue1, cue2 → cue1progress(진척)가 이루어지고 있는지 명확히 하기 위해 data(데이터)를 수집해야 함.data 항목:cue 전달 후 주저함이 없는지(no hesitation)behaviour가 어떻게 수행되는지consistent)it flows - no stuttering)strength in understanding)flexible)Performance(수행)는 실제 생활, 작업, 스포츠 등 최종 목표에 맞춰 behaviour를 발전시키는 과정임.behaviour의 질을 떨어뜨리지 않으면서 가변적인 환경에서 반복을 지속함.selective discrimination(선택적 변별)을 가르침.behaviour 간의 전환(transitions)을 더 어렵게 하여 난이도를 높임.minimised cues(최소화된 신호)를 도입함.behaviour를 더 길게 유지하는 것은 종종 동일한 behaviour의 반복임.session(세션)에 걸친 반복(학습 세션은 동일하거나 다른 학습을 포함하여 한 레슨 안에 여러 개 있을 수 있음)은 항상 progress를 보여야 함.progress가 없는 경우 그 원인을 조사해야 함.behaviour가 안정적인가:behaviour에서 요구되는 수준으로 정확한가(accurate).behaviour가 유연한가(단, 변덕스럽지는 않아야 함 - not variable).behaviour 덩어리보다는, 완성된 behaviour의 작은 조각(small slice)을 목표로 할 것.반복 훈련 중에 나타나야 할 행동의 6가지 변화 요소는 다음과 같습니다.
Is My Training Clean? CHECK LIST Your dog will give you the information you need to know whether your training is clean, quite mucky or a little dirty around the edges. Browse through, open the sections and check your understanding, skills and application. The training process goes backwards and forwards, to and fro between ourselves and the dog like a balanced conversation. Person The responsibility of the person, the knowledge and understanding of the technology. The skills they will need to practise and to communicate effectively. Technology This exists whether we like or don't like it, understand it or find it confusing. This does not control what we do, but it underpins how what we do, works. Dog Dogs do what dogs do, their behaviour is shaped by what works and what doesn't work. They seek information of value to them. Reward Process Planning Before the teaching begins: a review of where we are an outline of what we are going to do what the learner can benefit from learning Have we considered the reward delivery process to compliment the behaviour. Preparation Sufficient rewards available. Tools checked and handy: targets, note pad. Environment conducive to learning, no phones, no interruptions, distractions. Skills Have we isolated and practised the skills we are going to use Collect reward without losing connection to the dog Deliver cleanly, without confusion. Tossing practise, pointing practise, placing practise. Is it rewarding All rewards are reinforcers, but not all reinforcers are rewards. Make sure: Pleasure for the dog? Reward choice What is the reward FOOD : how large, what content, what size if to be thrown, how much scent is likely. PLAY: have we taught the appropriate rules, aware of the safety requirements and the Start / Stop cues. AFFECTION: tactile contact, has this been proven to be reinforcing in this environment for this behaviour. Where stored Can the location of the reward reserve affect the dog or influence the behaviour? Are we wearing the food, carrying the toy or do we go to collect it? Are there multiple locations? Where delivered Does the placement of the delivery set up the start of the next behaviour? How delivered Does the delivery pattern affect the repetitions of the next behaviour? Does a consistent pattern contribute specific anticipation into the behaviour? Does the pace of the delivery affect the behaviour? Does the pace suit the speed of the learner? Familiarity Is the dog familiar with the reward pattern? Can they recognise the cues for different delivery patterns? Is the dog familiar with the difference betwe
en take from hand and follow hand when luring? Is the dog clear when they go to collect the thrown food? Is the dog clear when you will go collect the food? Is it motivating? Does the dog have a strong interest in the available rewards? Do we have good evidence of that? Could the anticipation cause over arousal? The purpose of learning is to earn more and better rewards Schultz Antecedents Learning space Have we set an environment to compliment the learning? Are we aware of changes in the environment? Have we planned the learning that enables us to clearly see what is happening? Do we need resting areas and learning areas to be separate? Cue Skills Have we planned the cues? Have we chosen appropriate words, tone, pace, pitch? Have we practised how we deliver the cues? Are we aware of pre-cues? If we are giving a signal, how is it viewed by the dog, what movement opens the signal, how long will it be presented for. Do we deliver the cues when the dog gives a cue: ie are they cue seeking. Are we disciplined to NOT repeat the cue. Transferring the cue Are we aware of the temporary cue during learning (which may be a target or a lure). Do we understand the procedures for transferring the new cues? Have we planned a performance cue for the life of the behaviour? Can we add a novel cue to an existing behaviour that already occurs. History What is their experience? What associations are in the past with these antecedents? The Environment This includes everything around us and the dog. Certain elements are more relevant to the dog than us, such as scent. We are also part of the environment and affected by the environment (if you know you are being videoed ... does it change your behaviour?) Pre-cue There are pre-cues that the dog will be looking for. This may vary from an intake of breath, to eye contact, a poise, a raise of eyebrows. We want pre-cues, we do not want cues launched without preparation. Pre-cues may be coming from the environment The Relevant Cue This is what we would like the dog to select and respond to. Emotional context Is the dog emotionally able to learn? Could the dog be over-aroused? Could the dog be over-whelmed? Can the dog focus? Is the dog interested? Does the dog feel safe and comfortable? Shall we begin Have we taught clear cues for the learning to begin (open session cues) Have we taught clear cues for the learner to take a break (temporary end the learning) Are we clear that the dog is telling us they are ready and when they are finished? Do we have a cue to call it a day and this not be a dissappointment. Environment History Does
the dog have associations with this environment that would conflict with learning? Has the dog a history and expectation of reward in this environment? Recognise Can the dog recognise the given cue(s)? Is the dog in a place to see the cues clearly? If the dog has seen the cue are we sure they can remember what behaviour it pairs with? Cue seeking Does the dog seek cues? Have we taught and rewarded cue seeking? Antecedent presented Dog Responds Pace The pace of the response is critical information about the ease of remembering, desire to repeat, understanding of what to repeat. This is often referred to as fluency, or latency. This is under the control of the dog. Where this response changes is an indicator of mental fatigue. Behaviours Planning to teach What behaviours are we engineering for the dog to learn? Is this constuctive and avoiding a suppression of behaviour? Aim to encourage what we want not what we don't want. Do not seek an absence of behaviour. Teach ethically appropriate for that individual learner. Is this behaviour going to change as the dog matures. How much Guidance should we give: read more Luring skills Are you familiar with using your hands, or a cup on a stick, to teach the dog how to follow a lure. Can you alter the lure pace and engagement to maintain following and avoid biting or loss of focus. Can you mark by changing the lure/follow hand into a take-it-now hand. Shaping skills Are you familiar and confident with the dog becoming self-directed in their learning? Can you avoid extinction as part of their progress? Can you build a repetition series to add strength before progress? Moulding skills Moulding behaviours through touch-prompting, using the environment to limit the responses. Assessing the difference between the environment proving information that is voluntarily sought, or pressure that is implicit. Targets Target acquisition: how is the behaviour, association or action learned? Using targets as prompts to teach new behaviours. Replacing the target. Base behaviour Connection is our base behaviour; it is where the learning begins. Connection requires you to be aware of your dog's needs, limitations and abilities. Connection requires you to recognise when fatigue is starting. Does the behaviour require: a consistent starting position, such as a sit or a down or a stand a consistent location, relative to you, perhaps in front facing you or at your side or at a distance. New When planing new learning: does this behaviour exist naturally for this individual what is the function for this behaviour in its natural state what i
s likely to be the reinforcer in that state What is the future of this behaviour in 6 months, in 2 years, in 5 years? Is this behaviour a component of a future behaviour Does this behaviour need component behaviours already in place? What reward will compliment this behaviour? Can you anticipate that this behaviour is going to occur? Will you be able to see this behaviour occur? Consistent Is the behaviour stable: it is consistent in movement, speed and pace it is accurate, as would be required in the finished behaviour is the behaviour flexible, (but not variable) Aim for a small slice of the finished behaviour rather than a pool of the whole behaviour that needs "tidying up later". Grow organically Behaviours should be allowed to develop from within the dog not by a veneer applied externally. Behaviour build on rewards will naturally gather strength. Fit for the task Is the dog physically fit and prepared for the movement or action? Is this behaviour a natural behaviour for this dog? Environmental support Does the environment support or conflict with this behaviour? Beneficial Is learning this behaviour going to be a direct benefit to this dog? Error response Have you planned your response to the unexpected response from the dog? What data can you obtain: was the cue given when the dog was cue seeking / ready to respond was the cue clear, consistent and unambiguous was the response a similar behaviour, behaviour that you have been practising recently Is the dog confident to offer the default when uncertain? Not good enough Is the response correct but of a poorer quality than previous shown? Is there a missing component or skill that can be refreshed? Could fatigue be starting? How will you respond? Default Have you taught a default behaviour which the dog should perform if they do not recognise the cue, or would prefer not to respond? The dog is correct The dog should not become aware of failing, or making an error. Their effort to recognise the cue, respond with the associated behaviour should not be questioned. If the expectation is incorrect (from the person) then the dog is not at fault. They are responding to the environment and are true to their learning history. Poor quality If the dog is not responding with the expected quality, or standard the learning gap should be examined. The dog should be given information that they are responding correctly but that a reward has not been earned. A break is recommended for revisiting the planning. Reward Predicting Stimulus Markers Can you anticipate Markers and "good timing" after synonymous. Good timing
is a result of being able to recognise exactly when the behaviour is going to happen. It comes from closely observing the behaviour for the preceding actions, such as a tail flick before lying down, or a head tip before a sit. Marker cues what? Your marker as a cue is learned by the dog. You need to be clear about the response to your marker cues: ending the behaviour or maitaining seeking the reinforcement or waiting for delivery grabbing the toy or waiting for the cue to take it Targets Target acquisition: how is the behaviour, association or action learned Using targets as prompts to teach new behaviours Replacing the target Effective Does the chosen marker produce the accuracy in the repetition Is the marker suitable for the behaviour - does the dog need to see the marker or hear the marker. Reasons to Use a Clicker Markers are cues A marker should be paired with reinforcement processes. You can choose to use different process to specific markers. The marker is a cue to orientate and prepare to begin the reinforcement process. Consistent Is the behaviour stable: it is consistent in movement, speed and pace it is accurate, as would be required in the finished behaviour is the behaviour flexible, (but not variable) Aim for a small slice of the finished behaviour rather than a pool of the whole behaviour that needs "tidying up later". Marker response Does the dog show a response to the marker and evidence that they desire our reward? If the dog responds but does NOT seek the reward then reinforcement is likely gained from the environment. If there has been no response to the marker there should be no reward delivery. Delivery cue Is there a clear cue, or pattern, that gives the dog information about the reward process? Is the food going to be breakfast in bed, thrown a catch or placed? Does the dog respond to this pattern cue? Variety of markers If a variety of markers are used is there evidence the dog is responding correctly to each marker? Does the different marker represent a different reward process? Reinforcement Process No cheer leading Offering anything except cues, prior to the marker (the start of the consequence) can hide the weaknesses in the behaviour. Cheerleading the dog to success will often require cheerleading as a cue for life. Once the marker has occurred we can cheer away with volume and sincerity. Rates of reinforcement A rate of reinforcement must be relevant to either: A) the number of successful behaviour cycles (Antecedent, Behaviour, Consequence) per minute B) the number of behaviours occurring Rates should only be measured fo
r each behaviour, not in comparison between behaviours and not between learners. Read more: Fast Does not Mean Better Cue seeks again After the reinforcement process completes, or consumption occurs, does the dog initiate cue seeking again? If there is an increased delay to the previous repetitions this may be an indication of mental fatigue. Repeat Cues to continue Continuation is a series of repetitions of the same behaviour Cues can be added to assist the maintenance and then faded from the rear, forwards. cue1, cue2, cue3, cue4 cue1, cue2, cue3, cue1, cue2, cue1 Measure progress Data should be gathered to be clear on what progress is being achieved. Data on: no hesitation after the cue is delivered how the behaviour is performed consistent it flows - no stuttering it has strength in understanding it is flexible Develop Performance "Performance" is developing the behaviour to its final goal, whether real life behaviours, work or sport. Continued repetition with variable environments without diminishing the quality of the behaviour. Teach selective discrimination in the presence of potential disruptions. Increase difficulty levels with harder transistions between behaviours. Introduce minimised cues. Article on fading in disruptions Repeat and duration Extending a behaviour for longer is often a repetition of the same behaviour. We give information what is required, not withhold information and hope for the best. Read more: Duration or sustaining a behaviour , Progress Repetition over several sessions ( a lesson can have more than one session, of either the same learning or different learning ) should always show progress. This should be monitored and a lack of progress should be investigated. Consistent Is the behaviour stable: it is consistent in movement, speed and pace it is accurate, as would be required in the finished behaviour is the behaviour flexible, (but not variable) Aim for a small slice of the finished behaviour rather than a pool of the whole behaviour that need "tidying up later". Visible progress During repetition the behaviour should show an increase in: familiarity: there should be a visual increase in pleasure in performing the behaviour flow: there should be no stuttering or hesitation during the behaviour strength: if there are physical components then there should be a visible increase in agility, ease of movement and physical power knowledge: we can never know what the dog knows, but the behaviour should look "full of understanding", where the dog show they know what to d and how to do it consistency: a consistent movement that
does not change in structure, but only in fluency flexible: a demonstration that the behaviour remains consistent when there are minor changes to the environment, such as distance to travel to commence, a change in surface etc.