Training and generalization of yes and no as mands in two autistic children.
Modeling plus play reinforcement teaches autistic kids to mand "yes/no," but nonverbal learners need practice across five item sets before the skill generalizes.
01Research in Context
What this study did
Two autistic kids, ages 6 and 8, got yes/no training. One spoke in short phrases. The other had almost no words.
The trainer showed a toy, then modeled "yes" or "no" while holding the item. Correct answers earned 30 s of play and praise. Wrong answers were ignored.
Training moved through five different toy sets to test generalization.
What they found
Both children learned to mand "yes" and "no" in about 10 sessions. They kept the skill one month later.
The verbal child generalized after one toy set. The nonverbal child needed all five sets before saying "yes/no" with new toys.
How this fits with other research
Leon et al. (2010) repeated the idea with attention mands. They also saw generalization, but only when teaching happened during real reading time. Together the studies show mand training works best in natural moments.
Stagnone et al. (2025) took the next step. They first ran a functional analysis to prove that repeated "what's that?" questions were really attention mands. Then they taught a simple "excuse me" mand and the questions dropped. Their work extends Adams (1980) by showing you should check why the child talks before you pick the mand to teach.
Lord et al. (1997) add urgency. They found more behavior problems in people who have no words. Teaching yes/no early, as Adams (1980) did, may prevent those problems.
Why it matters
Yes/no is power. A child can accept or reject any item, activity, or break. Start with modeled trials and quick play reinforcement. If the child is nonverbal, plan to practice across at least five item sets so the skill travels. Run a brief functional analysis first if the child already uses odd phrases; the real function may be attention, not information.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick five small toy sets, model "yes/no" while holding each item, and give 30 s play for correct responses.
02At a glance
03Original abstract
This study presented a practical and replicable procedure to train and generalize the use of "yes" and "no" as mands by a nonverbal boy and a previously echolalic girl, both diagnosed as autistic. The procedure used systematic modeling and reinforcement with detailed criterion for introducing and terminating the training stimuli. The subjects were first trained to use yes and no to mand three food items, following the stimulus question "Do you want...?" and presentation of a food item. The teachers then tested the subjects for generalization of the two mands to successive sets of new food items. The results showed that the nonverbal subject needed to be trained on five sets of food items before generalization occurred. The previously echolalic subject generalized and maintained the two mands after being trained on only one set of items. The study thus demonstrated that the procedure was effective in training two useful mands for the autistic children, and that after such training, the behavior may then generalize to new items without training. Specific response patterns and the importance of intermittent modeling and arrangement of reinforcers in the training are discussed.
Journal of autism and developmental disorders, 1980 · doi:10.1007/BF02408465