Evidence for the application of rules in Pavlovian electrodermal conditioning with humans.

Two Pavlovian SCR conditioning experiments investigated interference effects in sequential training of positive and negative patterning discriminations in humans. In Experiment 1, positive patterning (A-, B-, AB+) was trained in Phase 1, immediately followed by a negative patterning schedule (C+, D+, CD-). We predicted that human participants would learn a specific numerosity rule in positive patterning, which interferes with the subsequent negative patterning schedule. In Experiment 2, negative patterning (C+, D+, CD-) was trained in Phase 1, followed by a positive patterning schedule (A-, B-, AB+) in Phase 2. Because human participants would learn an abstract 'separate-versus-together'- or 'opposite'-rule to solve the negative patterning discrimination in Phase 1, there should be less interference in positive patterning in Phase 2 where the separate/together-rule could be applied, too. In both experiments, the initial patterning discriminations were acquired successfully. In Experiment 1, human participants totally failed to solve the Phase 2 discrimination, while in Experiment 2 appropriate response differentiation developed in Phase 2. Thus, without pre-experience human participants seem to utilize a specific numerosity-rule in positive patterning and a separate/together-rule in negative patterning.