Deriving relations at multiple levels of complexity following minimal instruction: A demonstration
A short burst of color-character-action drills created full derived networks that kids used on their own in Candyland.
01Research in Context
What this study did
Four kids with strong language skills sat at a table.
The experimenter ran short discrete-trial lessons.
Kids learned to match color cards to cartoon heroes and to actions like jump or spin.
No pictures from the board game Candyland were used during training.
After only a few minutes, the team tested if the children could mix and match the new links without being taught.
They also watched to see if the kids would use the links later when they played a real Candyland game.
What they found
Every child showed emergent relations.
They could pick same, opposite, and combined links they had never practiced.
Later, when the Candyland spinner landed on a color, the kids smoothly named the matching hero and acted out the action.
The brief, table-only drills had built a network that transferred to the game.
How this fits with other research
Zhirnova et al. (2025) also built networks in typical kids, but they added naming and listener tests to create analogies.
Paranczak kept instruction even shorter and still got full networks, showing minimal trials can be enough once language is solid.
Luoma et al. (2024) worked with adults and had to add extra tact and intraverbal training for comparative relations to emerge.
The adult study looks like it clashes with the child study, but the difference is age and task type.
Kids grasp simple coordination and opposition faster, while adults need more help with subtle comparisons like bigger or smaller.
Nakagawa (2005) and Brown et al. (1994) proved that rats and pigeons can form equivalence too, so the process is basic across species.
Paranczak closes the loop by showing the same rules work in a fun, child-friendly game.
Why it matters
You can build flexible language and play skills with only a few tabletop trials.
Teach the core relations first, then drop the child into the natural activity.
The game itself becomes both reinforcer and probe, saving you hours of drill.
Try this when you need quick generalization from table to play with verbal clients.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick three game pieces and teach two matches each in DTI, then watch the child use the untrained links during the actual board game.
02At a glance
03Original abstract
Recommendations for achieving generalized instructional outcomes often overlook the capacity for generative learning for most verbally competent humans. Four children (ages 5-8) participated in this project. In Study 1, we provided decontextualized discrete trial teaching to establish arbitrary relations between colors, pictures of characters, and researcher motor actions. All participants engaged in derivative responding, providing evidence of relational framing. Subsequently, we demonstrated that, with no additional instruction, these derivatives contributed to effective action within a socially valid context (i.e., Candyland gameplay). Study 2 extended the demonstration by teaching frames of opposition. Following teaching, all participants engaged in novel and contextually appropriate responding that entailed the derivation of both coordination and opposition between untrained stimuli. This outcome demonstrates how teaching simple relations can result in learning that manifests at higher levels of complexity (i.e., relational networking), providing some evidence that there can be socially valid benefits to decontextualized discrete trial instruction.
Journal of Applied Behavior Analysis, 2024 · doi:10.1002/jaba.1067