Five hundred msec after the fixation point vanished, a pair of digits appeared and remained visible until the participant responded or for 5,000 msec. The next trial began 1,000 msec after the disappearance of the stimulus. Data collection and stimuli presentation

were controlled by a Compaq computer with an Intel Pentium III central processor. Stimuli were presented on a Compaq S510 monitor. Participants sat approximately 60 cm from the computer screen. A QWERTY keyboard was placed on a table between them and the monitor, and they were asked to respond www.selleckchem.com/products/XL184.html manually by pressing the key attributed to the numerically larger digit. In the horizontal version, the participants were instructed to press a left key (“”F”") if the left digit was larger, and to press a right key (“”J”") if the right digit MEK phosphorylation was larger. In the vertical version, the participants were instructed to press a bottom key (“”B”") if the bottom digit was larger, and to press a top key (“”Y”") if the top digit

was larger. To avoid a possible artifact in the vertical block, all participants were asked to use their right index finger for the top key and the left index finger for the bottom key. Mean RTs of correct responses were calculated for each participant in each condition for the numerical and physical comparisons, separately. These mean values were subjected to 3-way analysis of variance (ANOVA), with physical-numerical congruency (congruent, neutral and incongruent), and number-line compatibility (compatible and incompatible) as within-subject factors and with group (synesthetes and controls) as a between-subject factor. Incorrect, very short (≤150 msec) or very long responses (≥2,000) were excluded from the

RT analysis. Mean RTs and ERs (error rates) in the various conditions are presented in Table 2. The results for the vertical presentation corresponded perfectly with our expectations. A significant main effect was found for dimension congruency [F (1, 15) = 57.5, MSE = 834, p < .0001]. That is, RTs for congruent trials were significantly faster than RTs for the neutral trials, which were significantly faster than RTs for the incongruent trials. Nearly significant effects were found for number-line compatibility [F Alanine-glyoxylate transaminase (1, 15) = 4.3, MSE = 1882, p = .05] as RTs for the compatible condition were faster than RTs for the incompatible condition. No other main effects or interactions were found; meaning the numerical comparison groups did not significantly differ in their patterns of behavior ( Fig. 1A). A significant main effect was found for dimension congruency [F (1, 15) = 19.2, MSE = 866, p < .0001]. The interaction between congruency and number-line compatibility was found significant as well [F (2, 30) = 13.5, MSE = 600, p < .0001]. Importantly, these two variables also interacted with group [F (2, 30) = 4, MSE = 600, p < .05].