Training in Systematic Instruction
Dr Mark Kilsby and Julie Allan
Mark Kilsby (PhD) December 2015
The methods of Training in Systematic Instruction (TSI) fit broadly into the in-work support element of the 'Supported Employment' approach. It was originally devised by Dr Marc Gold, an American Psychologist who conducted his pioneering research in the mid 1970’s and early 80’s. It was first introduced into the UK in the late 1980’s. I have been delivering TSI workshops consistently in the UK since 1991 and have presented alongside a number of home grown contributors. The views expressed below are based upon my own experiences and contributions made in the delivery of TSI.
TSI is an approach intended for those working directly with people who require additional support to learn the practical skills involved in work and independent living. Originally, it was conceived with those with learning disabilities in mind, but its relevance and effectiveness to reach across many vulnerable groups has become obvious, including to people with autism, Asperger’s syndrome, users of mental health services and basically anyone who lacks confidence and/or needs support to successfully undertake and learn the practical tasks involved in work.
When TSI arrived in th UK it made an immediate impact as it held the promise of providing an effective method for training people with learning disabilities complex work skills which up to that time, were thought by most in society, to be beyond their capabilities. Not surprisingly, it became an instant hit with those of us involved in setting up or developing supported employment services in the UK since it offered a vehicle to support people with learning disabilities into inclusive paid work environments.
Focus on Practical Tasks
The model task used in TSI was and often still is, the bicycle brake – not an ordinary calliper brake, but a complicated hub brake, beloved to BMX bike riders. The bike brake, used as a 'model' task' has some very positive properties which make it excellent for TSI purposes - it can fit on a desk, provides a complicated challenge to the course participants; and maybe most importantly, it has a focus on 'practical’ rather than 'academic' skills. This allows us, as TSI advocates, to demonstrate that whereas, many service users do not fare well on academic tests, they can and do perform very well on practical ones.
This does not preclude the potential of TSI to be used for teaching academic skills, but there are fundamental differences between teaching academic and practical skills. Firstly, although it may be ok for a pass mark in an academic test to be set at 65%, this does not work for practical tasks. Practical tasks need to be 100% correct; you are hardly going to congratulate your car mechanic for setting your brakes at 65% working efficiency!
What is TSI?
Pioneer of the "Try Another Way" Approach
Dr Marc Gold
Our TSI Course is placed within the framework associated with the Values of 'Social Role Valorisation' as devised by the late Dr Wolf Wolfensberger and his associates. link here
Dr Wolf Wolfensberger
Pioneer of 'Social Role Valorisation' Theory
Readers are encouraged to explore the various websites dedicated to Marc Gold, while being mindful that the approach has developed and changed substantially over the years.
Second, without going into the minutia of reward/motivation systems, practical tasks require a different method of appraisal than is commonly used for academic tasks: in academic appraisal, feedback is more or less provided at each step of the thought process, including the conclusions/answers provided. If the same approach is applied to practical tasks (e.g., saying "that's right" or "well done" following each of the 24 steps of the bike brake), then there is a real danger that the trainee will become dependant upon trainer rewards, rather than the intrinsic rewards for successfully undertaking and learning the tasks independently.
Train don't Test!
Most TSI courses involve the use of what we describe as 'volunteer helpers'. These are a group of people (usually with a learning disability) who join us on the course on the 3rd day: the challenge for the participants is then to apply the methods of TSI and train (one to one) the helpers to assemble the bike brake. Marc Gold once said that he could find no correlation in his data between people’s IQs and their ability to perform practical tasks. We would absolutely concur with this: we’ve found absolutely no correlation between peoples’ academic abilities (course participants and the disabled helpers) and their ability to do practical tasks indeed, in our experience very often the trainees learn the task faster than the trainers!
We can almost guarantee that of the people with learning disabilities who help us on the course, at least 30% learn to assemble the bike brake faster than those attending the course as trainers. In one of our courses we observed all the trainees learned the task faster than the 10 course participants! This is something which those charged with assessing people’s abilities should take extra special note of - even after all these years we still find this remarkable and its uplifting and incredibly potent.It also has obvious implications for realising a greater potential for people and countering the barriers which exist due to associated with the Competence/Deviance Hypothesis.
This prompts questions about how we are assessing people’s abilities. For example, the potential for people with learning disabilities to access apprenticeships is severely hampered by the prerequisite requirement of level 2 English and Maths. While many with learning disabilities may well have special educational needs, their ultimate potential to prosper in the world of work will lie in their practical abilities and the school and college curricula for those with so called ‘special needs’ should reflect this fact far more than is currently the case.
The Problem of 'Learned Helplessness'
Another context for consideration when developing TSI has been the challenges faced by many vulnerable due to learned helplessness. We can trace the lack of confidence shown by many people with learning disabilities when they attempt tasks to a history of failure, negativity and low or no expectations, some of which originates from continually being ‘punished’ for making the wrong decisions. By punishment, of course, we certainly don’t mean biffing the trainee around the head, punishment can occur in much subtler ways for instance:
During our workshop, one of my roles (cast as villain) is to politely ask one of the course participants to pick up a cup of water strategically placed to ensure they have to lean forward and reach to get it. As soon as they lean to pick it up but before they reach it I sternly say “NO”! Usually the participant recoils back into their seat looking rather confused and a little shocked. They are then politely asked again and again to pick up the cup, each time to be greeted with a shout of “NO”! Predictably, most people stop reaching for the cup before their 4th attempt and also, unsurprisingly we get a few expletives from some of those participants!
The problem here is that simply saying ‘no’ imparted no information other than telling the person ‘they are wrong’. I effectively ‘punished’ the person for making a decision! As readers with only a rudimentary knowledge of behavioural principles will know, this vastly reduces the tendency for people to make their own decisions. Because of this, in TSI the focus on what in behavioural science speak are rather coldly called ‘antecedent cues’. These are prompts provided by the trainer to the trainee with a disability (such as pointing to the next bit of a task) that occur before the learner has made a decision. These cues are potentially dangerous, because if they are used too often they have the potential to take the trainee out of the decision making process altogether and can contribute to what is often termed ‘learned helplessness’.
All this might sound like common sense, but it is also counter intuitive. One problem that presents itself to the trainer due to learned helplessness is that since the trainee is hesitant to make decisions, the trainer is tempted to provide an antecedent prompt (e.g., “that one next” – while pointing to the next bit). Throw in the tendency described earlier for some people to look for constant reinforcement and approval and you have a situation where many people will neither initiate nor move on in tasks unless there is someone to tell them what to do, when to do it and tell them how they have done once they have done it! It seems natural and kind to help people by telling them what to do next, ‘if the trainee wants help, give it em!’ In fact, of course this only compounds the learned helplessness people with learning disabilities experience, the trainer in this instance is not only in danger of becoming part of the problem but also the creator and compounder of the problem, creating a downward spiral of learned helplessness. In this respect, the term ‘taught helplessness’ is as least as accurate as the term 'learned helplessness'.
Over 25 years of delivering TSI and having watched literally thousands of people with learning disabilities assembling bike brakes; we can observe that the prevalence of learned helplessness, or vulnerability to it, remains high. For some people with learning disabilities it is severe; time and again we’ve seen trainees go to extreme lengths to draw the trainer into telling them what to do. It’s not unusual for some trainees to speedily learn the skills involved (there are 24 steps in our bike brake TA), yet make no decisions (initiations) throughout. One of the main advantages of delivering TSI is that you can demonstrate that this is 'taught' to the participants by training them to reverse this spiral: Providing the rules of TSI delivery are adhered to, you can literally remove years of institutional garbage in the space of 40 minutes.
Preparation and Training Strategies
Although there are many aspects to the TSI approach, there are two fundamental components that underpin its delivery. These are 1) preparation for training and 2) The delivery of a systematic approach to providing instructions and withdrawing training support efficiently.
As Benjamin Franklin once said “by failing to prepare, you are preparing to fail"
The first main element prepares the trainers to concentrate on the steps of the tasks to be learned. Clearly, you cannot train someone to do a task if you cannot do the task yourself! in this case the person invariably fails which can reinforce the false assumption that the person is unable to complete the task, when clearly it is a failure on the part of the trainer.
Preparation therefore, involves the trainer learning the task to be trained from an ‘expert’ (i.e. someone already doing the job) until they can do it blindfolded and then listing the sequence of steps involved onto a Task Analysis (TA) sheet - a rather tedious but very necessary part of preparing yourself to train. In recent courses, this has evolved into identifying both the action for each step and listing the conditions that arise due to that action.
Importantly, ‘criterion’ (i.e. what constitutes the correct action) are specified for each step of a task. The reason behind this is that in looking ahead to training (distinct from preparation) it is the trainer’s responsibility to ensure that at each step of the task the trainee (with or without assistance), reaches step criterion - which for those of you familiar with the approach is described as being ‘Topographically Correct’.
There are a couple of reasons for insisting on this: Firstly, it would be no good if I trained a trainee to make a cup of tea in the morning and then in the afternoon another trainer came in and trained them in a completely different way - we’d have one very confused trainee. By specifying each correct action, the TA maintains consistency, providing the trainer with a method for specifying the same set of actions during each training session with the result that any ambiguity around what constitutes being ‘correct’ is minimised, thus ensuring the task is trained in exactly the same way over successive training trials; no matter who trains it, or on what occasion.
Secondly, specifying the exact action for each step reduces the likelihood of potentially stressful errors occurring. Therefore during training, the timing of assistance is crucial and the trainer is encouraged to intervene before an error is made on any particular step by the trainee. - If the amount of errors are reduced, so too is the time it takes the trainee to learn it and crucially, without compromising safety. Importantly, because the TA is based on the ‘natural method’ (i.e., it’s done in the same way as experts who already do the job) it does not make the trainee stand out as being different which goes a long way to helping in the development of those all important natural supports and countering the issues associated with the Competence/Deviance Hypothesis.
Amount of Instruction - Applying the Prompting Hierarchy
The second main element involves the way we actually assist the person (i.e. train them). This involves providing assistance (through demonstrations and physical, gestural and/or verbal instructions) by using ‘prompting hierarchies’, which is a posh way of saying that when assistance becomes necessary, the trainer should start by providing the least ‘powerful’ forms of assistance and only increase the amount of support provided if it becomes necessary to do so.
So, instead of starting with a powerful physical assist, the trainer attempts to try a gesture first, if the gesture works - all well and good but if it doesn’t, the trainer should ‘up the power’ (i.e. increase the level of support) immediately to a physical prompt or risk flapping about with gestures that resemble a game of charades rather than instructions. As a general rule, when people first start using the approach, they have a tendency to provide too little ‘power’, perhaps due to an over protective reaction from within the human service culture which has created in some of its staff, enormous hesitancy to touch people; even if that physical contact has a teaching purpose.
Using prompting hierarchies in this way has 3 main advantages: Firstly, starting with the least amount of assistance disciplines the trainer to naturally decrease or fade the amount of assistance they provide on the basis that the less the trainer is doing the more the trainee is doing. Secondly, it does not de-skill the person with too much assistance and finally, it makes the approach applicable to people with a wide range of disabilities. When TSI is delivered properly, it can be applied to those with the highest levels of support as well as those requiring less intense levels of support.
In other words, you gauge the amount of support needed by the trainee, if people need lots of support you provide it; where they need less you provide less, but you have to know how, when and what type of support to provide when it is required. And that’s the art in the science!
Emphasis on Self Determination - Timing of Assistance
Although there may have been many changes to TSI over the years, perhaps the biggest one has been a focus on the initiation of independent decisions made by the trainee about what step or routine of the task comes next. This is distinct from ‘skill acquisition’ - which is the trainee’s ability to perform only the physical aspects of the task.
This dichotomy leads to a switch the emphasis, away from skill acquisition, towards the decision making part of the process. The reasoning behind this is that over time, some of us delivering the course came to believe strongly that people’s abilities and confidence to make these decisions was more important, especially during the early part of training, than being able to actually carry out the skills involved in a task. In other words: it matters little that someone has the skills to physically undertake a task if they are not able or confident enough to make decisions on their own pertaining to its successful undertaking!
Put that antecedent finger away – It’s rude to point!!!
Many of you who have taken a TSI workshop will recall the agony of trying not to point to the next piece in the bike brake task, when a trainee is constantly asking verbally or non-verbally “what’s next?, what’s next?, is it that one?” The problem of course is not just that in this situation the trainee is often looking to draw you into some form of prompting, but also that some trainers literally have what we term an ‘antecedent index finger’. Ah yes there it is time and time again, although not actively pointing to the next bit, it’s ready and prepared for action at the first opportunity; sitting on the desktop, twitching and pre-programed to take on an antecedent mind of its own!
This is not to say that as a trainer you should never use antecedent cues we’ve encountered numerous situations, where on the first training trial of the bike brake (in our course the trainees go through the task three times following a demonstration), the trainee has not initiated one decision on their own. The challenge, as many of you will recall is to somehow turn this around. Crucially, once the person has made a decision, even if it is wrong, we avoid saying ‘NO’, which for reasons discussed earlier could stop decision making in its tracks. It’s also important to point out that saying “no” is not necessarily always verbal; it could be physically stopping the person’s hand or worse, a finger wag, which we really have seen on occasions.
Instead, as trainers we provide corrective feedback (e.g., pointing to the correct next step) after the person makes a decision but before they go on to make what could be a potentially punishing error. This is no longer an antecedent cue, but a ‘simultaneous’ one as it allows the person to make the decision prior to assistance being offered, but before an error occurs. This corrective feedback avoids simply telling someone they are wrong and provides them with the information they need to be correct.
The Concept of “No News is Good News”
One of Marc Gold’s most compelling contributions to TSI is the “No News is Good News” principle. This is based on the fact that many people, especially those lacking confidence and with a long history of failure, seek constant reassurance that what they are doing is correct. The problem is that this can result in learned (although you could say 'taught') helplessness with the trainee incapable of moving on from one step of a task to another without the trainers say so or reinforcement. In real employment constant reinforcement does not work because there’s not a boss in the land who will follow around an employee constantly saying “well done” or giving them a reinforcing nod every time they get something right.
In essence, the trainer must not become the reward for the person doing the job; the reward needs to come from completing the task independently and without constant prompting. Many of you will remember holding back any reinforcement from the trainees and only delivering it on the full completion of the bike brake task: with ‘no news is good news’ as a trainee you know you’re doing ok when the trainer is doing nothing. This sounds logical, but it is always a contentious part of the course, with many claiming that it goes against ‘everything they have ever been taught around teaching’. This is because teaching academic concepts and practical tasks are different: teaching academic skills involves more consistent reward systems based on feedback to the academic problems and challenges set by the curriculum through the assessment and marking procedures, whereas many practical skills are taken for granted.
The focus on decision making in TSI has also enabled us to demonstrate to course participants that even those who appear to be lost to the disempowering stranglehold of learned helplessness can actually be rescued and empowered relatively quickly; provided certain rules are applied. In all the years of running TSI course we have only once encountered a trainee who throughout the course of three training trials was unable to initiate a single action. Instead we have seen many more trainees with severe and complex needs who have moved from being antecedent cue dependant slaves, to initiating their own decisions over the course of only 40 minutes or so of TSI training.
This is very powerful, especially when as a trainer you’ve battled to stifle your need to reinforce after every step and hold back that pointing finger until finally the trainee initiates their first independent decision (albeit maybe even the wrong one), with only 5 steps of the final bike brake remaining. It’s an initiation and the breakthrough you’ve been working so hard for! It’s a very real and significant accomplishment for both the trainer and the trainee. Even now after 25 years I can still get a kick from that moment and seeing the look on most of the new trainer’s faces – pure, unadulterated joy and mutual reward: and best of all the trainer and trainee are working as a team! Once the trainees know they are not going to be punished for making a wrong decision, their confidence to act increases noticeably, reflecting no more than we probably already suspected: that people are more motivated and engaged when they are involved in making their own decisions! Once trainees have the confidence to initiate, they will work with you on acquiring the skills.
'Working as Equals'
Encouraging trainees to initiate is just one way to tackle the lack of confidence exhibited by many people with learning disabilities when confronted with a task. Often, as many will know, it manifests as self-doubt on the part of the trainees. On being confronted by the bike brake task many trainees (and many course participants for that matter) take one look at it and say two words: ‘Im Possible!’ In such situations it becomes even more crucial to avoid potentially punishing errors during training and as much as possible, the trainer needs to ease the pressure being felt by the trainees. It’s one thing to say this but quite another thing to do, after all, in our TSI course as a participant you are asked to: walk into a room with a person with a learning disability, who you’ve probably never met before and train them to learn a task, that only the day before you thought you might not be able to do yourself and to try out a training method you’ve never tried before, and then to cap it all off - we stick a video camera in your face and record the whole shebang!
So, the trainer is under pressure too and what is required is a dual approach, not one that heaps all the pressure onto the trainer. The trainer and the trainee need to be working together toward the same goal. One method that TSI has developed for taking the pressure off both the trainer and the trainee is to insist that the trainer overtly accepts responsibility when significant errors occur e.g., the bike brake falls apart at ‘the turn’ – a notoriously intricate step in the assembly of the bike brake. Such errors can and more often do, undermine the already fragile confidence of many trainees who will often throw their arms in the air and proclaim “See I told you I couldn’t do it”! In fact, if the trainee commits an error in this approach it’s the trainers fault, because the TA specifies the exact criterion that the learner must achieve in order to proceed to the next step and the error occurred because you (as the trainer) did not insist upon this. In such situations, why should the trainee take on any of the burden of responsibility? It was clearly the trainers fault!
In TSI, we therefore insist that when mistakes occur during the training session the trainer takes full responsibility by apologising(e.g., “sorry that was my fault”) and not punishing the trainee further by making them re-do it. Instead, the trainer should set it up from where the training broke down and the next time ensure they provide the appropriate support to avoid the error reoccurring. Most participants on our course get very nervous prior to undertaking the training with the trainees and it is surprising how often the trainees have provided them with encouraging support during training.
Increasing Societal Expectations
Taking TSI Forward
The majority of TSI courses have been delivered with supported employment providers in mind. Therefore, the framework within which we have delivered TSI has largely been in accordance with the supported employment values, aims and principles (e.g., Place-Train-Maintain, Zero Reject Philosophy etc…) with the explicit aim of enabling people with disabilities to progress or get direct access to inclusive paid employment.
Over the years we have learned to adapt the approach to meet different sector practices and contexts. For example with the current emphasis on transition from school to work and early intervention, Schools and Colleges and professionals, are engaging more with supported employment thinking in an attempt to create a more efficient link between their vocational courses and paid work outcomes. This requires a switch in emphasis towards more practically oriented teaching. Understanding the concepts that underpin TSI should, in my opinion, be a prerequisite for special needs teachers and tutors.
The relevance of TSI in training self-help and independent living skills also now seems obvious. For example, the skills involved in making ones own bed are very similar to the skills required for making beds in a hotel for a living. Although we have begun to deliver TSI training within the context of residential services, as yet we have barely scratched the surface in this area.
Developing an early intervention approach specifically for young parents also has to be a future priority for TSI and TSI leaders since those parents ultimately shape the lives of their sons and daughters based on what they as parents believe and know to be possible. TSI is a powerful tool to illustrate just what is possible and to increase parental expectations and their hopes and dreams for children with learning disabilities. Services, Colleges and Schools for people with learning disabilities will ultimately have to raise their game to meet those increased parental expectations – and as far as we can see, that can only be a good thing!