Article Figures & Data
Figures
- Figure 1.
Example of stimuli and the numerical Stroop task. A, Everyday digits and their corresponding artificial digits that were used to create the new numerical system. Each everyday digit appears above its corresponding artificial digit. Over 6 d of training, participants were instructed to refer to the artificial digits as representing various magnitudes and to decide on each learning trial which one of the two stimuli has a larger magnitude, while receiving visual feedback for their decision. Only adjacent pairs have been presented during this phase. B, An example for congruent and incongruent trials with artificial digits from the numerical Stroop task. In the current example, the symbols corresponding to the Arabic digits 3 and 8 appear on the left and right sides of a congruent and an incongruent trial, respectively. Participants were asked to compare the stimuli according to their physical size while ignoring their numerical meaning. On congruent trials, the larger number appeared in larger font, whereas on incongruent trials the larger number appeared in smaller font. The numerical Stroop effect (incongruent vs congruent) indicates the slowing in the decision time when numbers are irrelevant to the task and are therefore processed automatically (Henik and Tzelgov, 1982). Nonadjacent pairs have been presented in this task to examine transitive inference from the learned material (Tzelgov et al., 2000).
- Figure 2.
Learning functions for the three groups. Top, PPC group. Middle, Sham group. Bottom, DLPFC group. The slope for the PPC group (top) is steeper than the other two (middle, bottom), indicating faster learning. Analysis of the learning rate gave statistical support to this result by showing a significant difference between the learning rate for each group that was best explained by a linear trend: PPC > sham > DLPFC (p < 0.006); 92% of the variance explained.
- Figure 3.
Double dissociation between learning rate and automaticity as a function of brain stimulation. Although the speed of learning has been the fastest for PPC stimulation and slowest for the DLPFC stimulation, the automatic processing of the learned material showed exactly the opposite pattern. A linear trend analysis (group [PPC < sham < DLPFC], and task [learning rate ≠ numerical Stroop effect, RTincongruent − RTcongruent]) explained 99% of the variance (p < 0.006). Data are mean ± SEM.
Tables
- Table 1.
Reaction time in the numerical Stroop task with artificial digits and everyday digits, as a function of TDCS group for congruent, neutral, and incongruent conditions
TDCS group Congruity Artificial digits (ms) Everyday digits (ms) Mean SEM Mean SEM PPC Congruent 462 45 383 17 Neutral 429 41 400 15 Incongruent 470 56 436 23 Sham Congruent 456 39 393 15 Neutral 457 35 400 13 Incongruent 474 48 428 20 DLPFC Congruent 476 49 406 19 Neutral 484 45 414 17 Incongruent 568 61 444 25 SEM refers to the standard error of the mean.
In this issue
