General case simulation instruction and the establishment and maintenance of work performance.

Timberlake et al. (1987) worked with adults who had severe intellectual disability. They wanted to see if short simulation lessons could teach real job skills.

Each adult practiced only two examples of every work task. The team then checked if the learners could do six brand-new versions on the job without help.

The two-example plan worked. Trainees did the six untrained tasks correctly and kept the skill after staff stepped back.

Performance held up in the actual workplace, not just in the training room.

Davison et al. (1984) tried a board-game version of the same idea. Skills moved to a pretend workshop but not always to the real one. W et al. fixed the gap by training in settings that looked like the real job site.

Spealman et al. (1978) used a clothing-selection puzzle with institutionalized women. Their gains lasted months, showing simulation can work. W et al. moved the method into community employment.

Richman et al. (2001) later flipped the model: they taught supervisors with the same two-exemplar rule, then let supervisors train staff. The general-case logic held up, proving the idea works for trainers too.

You can save hours of training time. Pick two clear examples that capture the range of the job, run rehearsal with feedback, then test in the real place. Adults with severe ID kept the skill without extra prompts, so you can fade support faster and move on to the next vocational goal.