Class-consistent Differential Reinforcement And Stimulus Class Formation In Pigeons.
Reinforce longer looks at the correct choice to spark equivalence relations without extra drills.
01Research in Context
What this study did
The team worked with pigeons in a lab. They used match-to-sample tasks. Each trial showed three shapes. One shape was the sample. The bird had to pick the matching shape from two choices.
The trick was class-consistent differential reinforcement. If the bird peeked at the correct shape longer, it got more grain. This extra reward taught the birds to look longer at the right choice. Training kept the same reward rule every time.
What they found
Most birds showed transitive-like relations. After learning A-B and B-C, they picked A-C without direct training. More training made their choices even stronger. The birds also began to look longer at the correct shape before pecking.
How this fits with other research
Saunders et al. (1988) saw no symmetry or transitivity with plain match-to-sample. The new study added extra reward for longer looks. That small change flipped the result from zero to clear transitive control.
Cook (2002) later showed pigeons can learn broad same-different rules when mapping stays consistent. Both papers say the same thing: keep the reward rule steady and the birds build bigger stimulus classes.
Snapper et al. (1969) first proved pigeons can do any conditional task. The 1999 study adds a tactic: reward what the bird does while it looks, not just the final peck.
Why it matters
You can borrow the looking-time trick with learners who rush choices. During equivalence training, reinforce longer looks at the correct comparison. Use extra praise or a bigger token. The child may then show emergent relations without added drills. Try it next session if symmetry tests are flat.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →During match-to-sample, give stronger praise when the learner gazes at the correct card for two seconds before touching it.
02At a glance
03Original abstract
Match‐to‐sample training clusters of A1 (sample): B1/B2 (comparisons), A2: B2/B1, B1: A1/A2, B2: A2/A1, B1: C1/C2, B2: C2/C1, C1: B1/B2, and C2: B2/B1 were presented to pigeons with class‐consistent differential reinforcement using two dissimilar types of food reinforcers. Distinctive class‐consistent response patterns occurred to the samples during the fixed‐ratio 5 sample observing response requirement. Subsequent tests, modeled from the equivalence class paradigm demonstrated the emergence (80% class consistent) of the transitive‐like A‐C and C‐A relations for 4 and 2 of 12 pigeons, respectively, and a strong trend (over 70%) for 7 and 6 others, respectively; the emergence of the reflexive‐like identity relation when the nonidentical comparison was from the other class; and the disruption of the trained within‐class relation with the addition of a reflexive comparison. After directional training of C1: D1/D2 and C2: D2/D1, tests indicated no emergence of the symmetric‐like D‐C relation or the composite D‐B and D‐A relations, but the B‐D and A‐D transitive‐like relation occurred with some pigeons. Off‐baseline training with class‐consistent differential reinforcement contingent on responding to the D stimuli alone produced distinctive responding and, in turn, a trend to D‐C symmetric‐like control in 4 of 12 pigeons, as well as a shift toward class‐consistent control on D‐B and D‐A test trials. Class‐consistent differential reinforcement that produced distinctive sample behavior promoted stimulus control relations like those that circumscribe equivalence class formation. Respondent—operant interactions permit an analysis of the possible enrollment of stimulus values of distinctive responding to the discriminative stimuli forming the stimulus classes via processes corresponding to naming in humans.
Journal of the experimental analysis of behavior, 1999 · doi:10.1901/jeab.1999.72-97