The myth of ballistic processing: evidence from Stroop's paradigm

Abstract

The Stroop effect is widely considered to be compelling evidence that skilled readers cannot prevent themselves from reading the irrelevant word or even delay such processing. In contrast, several reports indicate that the Stroop effect can be eliminated by various simple manipulations. These reports have been criticized on several grounds, among them that the baseline condition is suspect. These criticisms are addressed by showing that when (1) a neutral baseline is replaced by congruent trials, (2) single letter cuing and coloring manipulations are combined, (3) attentional window conditions are blocked, and (4) the congruent/incongruent trial ratio is 20/80, the Stroop effect is eliminated. A second major finding is that despite no Stroop effect, negative priming is observed, consistent with the hypothesis that a distinct but delayed perceptual act processes the word. The default set may be to process to the highest level (semantics), but these reading processes are (contextually) controlled rather than ballistic.