It was demonstrated that central processing uncertainty (Hc) can be derived to provide a single valued statement of the information hypothesis RT=a+b (Hc) across test stimulus sets and across several levels of test stimulus probability in an information reduction task. The derivation procedure assumes successive tests of stimulus hypotheses with Bayesian revision of stimulus probabilities after failure of an initial test. It was shown that the procedure can be generalized to data from single test stimuli in an information conservation task. Stimulus and response repetition effects were estimated for the information reduction task data.