Programming "loose training" as a strategy to facilitate language generalization.
Let kids choose when to use new is/are words during mixed class work and the grammar shows up later on the playground.
01Research in Context
What this study did
Two students with intellectual disability practiced saying is/are sentences during regular class work.
The teacher mixed up the tasks. Kids chose when to talk. This is called loose training.
A multiple-baseline design tracked new sentence types and later play-time talk.
What they found
Both kids learned three new is/are forms.
Later, they used the words on their own during free play. The skill stuck for weeks.
How this fits with other research
Rogers-Warren et al. (1976) did an earlier cousin study. They rewarded preschoolers only when words matched real sharing. Both papers show kids learn best when adults loosen control and let the child lead.
Saunders et al. (2005) found that stronger expressive language links to higher development in autism. The 1982 loose-training method gives a practical way to build that expressive power.
Leezenbaum et al. (2019) warn that expressive-dominant toddlers with autism often have lower non-verbal scores. Loose training’s child-led style may fit these learners better than heavy drills.
Why it matters
You can weave grammar goals into everyday class jobs. Let the learner pick the moment to speak. The words travel to new places and last with no extra drills. Try it next time you run centers or art.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →During the next center activity, wait for the student to self-start a sentence with is/are, then praise; track if it pops up later in free play.
02At a glance
03Original abstract
This study investigated the generalization of spontaneous complex language behavior across a nontraining setting and the durability of generalization as a result of programming and "loose training" strategy. A within-subject, across-behaviors multiple-baseline design was used to examine the performance of two moderately retarded students in the use of is/are across three syntactic structures (i.e., "wh" questions, "yes/no" reversal questions, and statements). The language training procedure used in this study represented a functional example of programming "loose training." The procedure involved conducting concurrent language training within the context of an academic training task, and establishing a functional reduction in stimulus control by permitting the student to initiate a language response based on a wide array of naturally occurring stimulus events. Concurrent probes were conducted in the free play setting to assess the immediate generalization and the durability of the language behaviors. The results demonstrated that "loose training" was effective in establishing a specific set of language responses with the participants of this investigation. Further, both students demonstrated spontaneous use of the language behavior in the free play generalization setting and a trend was clearly evident for generalization to continue across time. Thus, the methods used appear to be successful for training the use of is/are in three syntactic structures.
Journal of applied behavior analysis, 1982 · doi:10.1901/jaba.1982.15-295