transfer of learning
You practised buttons on a practice board in the clinic and got quite good — but can you button your own shirt at home, in a hurry, before work? That jump, from what you trained to what you actually need to do, is transfer of learning. It is the whole point of rehabilitation: nobody wants to be excellent at clinic exercises; they want to live their life.
Transfer is the degree to which practising one task or in one setting carries over to a different task, situation, or environment. It tends to be strongest when practice closely resembles the real goal — similar movements, similar context, similar demands — which is the principle of specificity. Practising on a contrived device in a quiet room may not transfer well to a cluttered kitchen with a real shirt and time pressure. This is why mismatched practice can produce a patient who shines in therapy but struggles the moment they go home, a frustrating and common gap.
Maximising transfer drives many modern rehabilitation choices: practising the real activity rather than an abstract substitute, training in varied and realistic conditions, and, late in learning, deliberately adding the distractions and dual-tasks of real life. It also tempers hype — a clever computer game or robot may improve scores on that game without improving real-world walking unless the training is designed to transfer. The honest test of any therapy is not how the patient does in the gym, but whether the gains show up in their daily life.
A man practises a smooth reach-and-grasp on a clinic peg board and masters it, yet at home he cannot pick up a slippery wet glass over the sink; the trained task did not transfer, so therapy is changed to practise with real objects in a real kitchen.
The closer practice matches real life, the more reliably skills transfer.
Improving at a practice task does not prove transfer. A common error is celebrating in-clinic gains that never reach the home or workplace; transfer has to be checked in the real setting, not assumed.