Figure 2. RTs are best predicted by joint probability following sigmoid function.

(a) Calculation of joint probability (JP) and conditional probability (CP). From the sequence performed by the volunteer, we built a set of transitions which represents the locations along present trials (X0) as well as the locations of one previous trials (X − 1). Moreover, additional previous locations (X − 2, X − 3, etc) could also be included in the constitution of the set. The probabilities were calculated from the relative frequency the transitions of the set. (b) RT as a function of CP and JP for Experiment 1. Sigmoid and linear functions were fit to the raw RT vs. probability distributions. The color of the data point indicate the sequence performed by the volunteers. The transition probabilities were calculated considering three previous elements. (c) Explained RT variance by JP and CP, using linear and sigmoid function in Experiment 1. The JP associated to sigmoid function explain RTs variance equally or better than the other combinations for each number of previous locations used to constitute the transitions. (d) Same as Fig. 2b, for Experiment 2. The data point color indicate the number of the session. (e) Same as Fig. 2c, for Experiment 2. The RTs were best described by the sigmoid function and the JP values.