Modulation of conflicts in the Stroop effect.

The purpose of the current study was to evaluate the unique contribution of task conflict, semantic conflict and response conflict to the Stroop effect and to test how these conflicts are modulated by manipulating the proportion of neutral trials, known to affect the magnitude of the Stroop effect. In the first experiment, we employed the two-to-one paradigm (De Houwer, 2003) while adding neutral illegible stimuli, and in the second experiment, we employed two colors and four word colors. In both experiments, we created four congruency conditions (neutral, congruent and two kind of incongruent conditions-those that include response conflict and those that do not), which allowed decomposing the Stroop effect into three orthogonal conflicts. In both experiments, we also manipulated the proportion of neutral trials. Task conflict was defined by the contrast between illegible neutrals and color words, semantic conflict by the contrast between congruent and incongruent stimuli, and response conflict by contrasting the two kinds of incongruent stimuli. Our results showed that all conflicts contributed to the Stroop effect. Task conflict and semantic conflict were modulated by the proportion of neutrals but response conflict was not. These findings imply that task conflict and semantic conflict are part of the control loop of the Stroop effect, as conceptualized by Botvinick et al.'s (2001) conflict monitoring model. There is no clear evidence of the response conflict being part of the loop. To complete the picture, we also analyzed the conflicts in the Stroop task using the traditional dependent contrasts approach and found the basic pattern of results was similar. Thus, the main advantage of the orthogonal comparisons approach is the possibility to estimate the unique contribution of the conflicts contributing to the Stroop effect and their modulation of the Stroop phenomenon.