After repeated pairings of CS1 and US, the agent salivates when the light comes on (conditioned response, CR).
This demonstrates that the association between the CS2 and the US was initially learned but in comparison to the stronger predictive value of the CS1 there is no conditioned response.
[citation needed] Maes and colleagues reported fifteen experiments that attempted to replicate the blocking effect.
[5] They argue that publication bias may have produced a false confidence in the robustness of the effect.
[6] However, Soto (2018) has questioned this conclusion arguing that they come as a consequence of the type of stimuli used in these studies, and shows how contemporary models of associative learning can predict these results on the basis of this observation.