Increased number of responses may account for reduced resurgence following serial training
Pack at least three functional responses into FCT—resurgence drops because of response variety, not the order you teach them.
01Research in Context
What this study did
The team taught kids three ways to ask for a break.
They compared two teaching orders: all-at-once or one-at-a-time.
Then they stopped giving breaks and watched if problem behavior came back.
What they found
Resurgence dropped only when kids had three ways to ask.
The order of teaching did not matter.
More response choices, not teaching order, blocked relapse.
How this fits with other research
Ghaemmaghami et al. (2018) showed you can shape one simple request into a fancy one without relapse. Diaz-Salvat adds: give three choices from the start and you may not need shaping.
Sullivan et al. (2020) saw that when one problem returns, other untreated problems pop up too. Diaz-Salvat gives a fix: load the child with multiple FCRs so the whole response class has less room to return.
Fontes et al. (2018) proved that punishing a new request can bring back old problems. Diaz-Salvat hints that a bigger toolbox of requests may reduce the need for punishment in the first place.
Why it matters
You can cut relapse by packing at least three ways to ask for the same reinforcer. Teach them in any order you like; just pile on the options. Next time you write an FCT plan, list three mands before you worry about fancy shaping.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add two new mands for the same break to your current FCT plan and probe for resurgence.
02At a glance
03Original abstract
Abrupt discontinuation of functional communication training can cause resurgence of challenging behavior. Teaching multiple alternative responses in sequence (serial training) may reduce resurgence, relative to teaching a single alternative. However, previous evaluations of serial training included a different number of response options across comparison conditions. In Experiment 1, we varied both training type (single and serial) and number of response options, and replicated previous findings showing that more resurgence occurred following single training relative to serial training. In Experiment 2, we varied the training type while holding the number of alternative responses constant and obtained no consistent differences in resurgence. In Experiment 3, we varied the number of alternative responses while holding training type constant (i.e., single). More resurgence occurred in the condition with fewer response options, suggesting that the number of available alternative responses, and not explicit serial training of alternatives, was critical to outcomes.
Journal of Applied Behavior Analysis, 2020 · doi:10.1002/jaba.686