The punishing effect from the negative corroborations led to a drop in such a response rather than to a change in either the key-pressing or the estimations that the corroborations were referring to; responses which remained stable.
Table 3 contains participants' data from arrangement A for both transfer tests and subsequent trials registering verbal estimations about the tone contingency, and also the corroboration trials run in extinction for stimulus classes 1 and 2.
Instead of receiving the last three corroboration trial-blocks, she had three additional estimation trial-blocks, except for these being run in extinction in which the tone contingency was not operating.
Comparison participant 12 did not received corroboration trials.
Immediately after the first positive corroboration to C2, the participant ceased corroborating for all class 2 stimuli; a pattern that stayed constant during the remaining trial-blocks.