Retrospective Revaluation in Mice: Demonstrating Backward Blocking in Delay Classical Conditioning

Abstract

The current study aims to demonstrate backward blocking in delay eyeblink conditioning in mice As well, the SLG and Kalman Filter models, computational approaches that are able to accurately describe paradigms with retrospective revaluation, are examined. 13 mice were administered a backward blocking paradigm. The first phase consisted of compound conditioning, pairing a tone and light (CST and CSL, respectively) with a corneal airpuff (US). Phase 2 consisted of reinforced trials with one stimulus (maintained, CSM), while the second was not reinforced (omitted, CSO). At the end of each phase a test session was administered (T1 and T2, respectively), which included a higher proportion of test trials to obtain an accurate estimate of the CS associations. A control condition was included to control for any extinction effects generated by the CS-only test trials (present for both stimuli during both phases). It was found that the decrease in predictive power (measured by the CR success rate proportion T2/T1) observed in CSO was significantly greater than in the control condition. This result suggests the presence of backward blocking in delay eyeblink conditioning. This is the first demonstration of backward blocking in delay classical conditioning. Further implications are discussed, including future experimentation to examine whether the cerebellum alone is able to sustain backward blocking as a form of retrospective revaluation.