Transfer of function across members of an equivalence class.
Once stimuli join an equivalence class, any response rule you teach to one member will appear in the others—so probe the whole class before extra training.
01Research in Context
What this study did
The team used a computer matching game. Participants learned to pick STAR when they saw TREE.
Later they learned to pick WORM when they saw STAR, and BLOCK when they saw TREE.
No one ever directly trained WORM-BLOCK, yet the four pictures became one family.
What they found
After the class formed, response rules trained to STAR and TREE jumped to WORM and BLOCK.
For example, if fast clicks earned points for STAR, fast clicks also showed up for WORM.
The function slid across all members without extra teaching.
How this fits with other research
Aman et al. (2002) extends this idea. They built classes by pairing shapes with mild shock. No matching games were used, yet the function still moved.
Fields et al. (2021) supersedes the old computer set-up. They separated the choice window from the pictures and doubled class formation success. Use their timing tweak to get the same transfer faster.
Pérez-González et al. (2003) adds context control. They show that background colors can also enter the class and steer new choices. Plan your room or screen colors with this in mind.
Why it matters
You now have three levers: matching, respondent pairing, and context cues. Pick the one that fits your client and setting. After class formation, probe all members for the target skill; you may find the learner already performs it without direct teaching.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →After a learner masters A→B and B→C, test the untaught C→A and watch the skill appear—then move on instead of reteaching.
02At a glance
03Original abstract
A child's presses on response windows behind which stimuli were presented via computer monitor occasionally lit lamps arranged in a column; a present was delivered when all lamps in the column were lit. During the operation of a multiple schedule, the child first learned low rates of pressing in the presence of STAR and high rates in the presence of TREE. Later, in an arbitrary matching task, the child learned to select STAR given wiggly WORM and TREE given BLOCK. When WORM and BLOCK were inserted into the multiple schedule, the low and high rates respectively correlated with STAR and TREE transferred to them. Tests of reflexivity (identity matching) and of symmetry of the arbitrary matching implied that STAR and WORM had become members of one equivalence class, and TREE and BLOCK had become members of another.
The Analysis of verbal behavior, 1989 · doi:10.1007/BF03392841