An evaluation of generalization of mands during functional communication training.

The team ran FCT with sign language for one reinforcer only.

They first taught the sign in that single context.

Later they tested if the child would still sign when the same reinforcer was now available in new, untrained situations.

The child kept signing in the new contexts even though no one had drilled him there.

Problem behavior stayed low while the sign carried over.

Neely et al. (2018) looked at 37 FCT papers and found most never checked if the skill lasted outside the teaching room. The current study is one of the few that did, so it fills a gap the review flagged.

Boyle et al. (2019) took the same stimulus-control idea but taught two signs for two reinforcers after mixing the functions together. They also saw clean stimulus control, showing the 2013 result holds when you make the task harder.

Tsami et al. (2020) sounds like a contradiction: only one of five kids kept the gains when conditions were pulled apart. The difference is they started with combined contingencies, while S et al. started narrow and then broadened. Starting small and building out may be the safer path.

You can save hours of extra teaching. Train the mand in one tight context, then probe new ones before you add more lessons. If the sign shows up and problem behavior stays low, you are done. If not, you know exactly where to add practice.