TRAINING ON INTEGRATED VERSUS SEPARATED STROOP TASKS - THE PROGRESSION OF INTERFERENCE AND FACILITATION

Two experiments examined the course of interference acid facilitation in the Stroop (1935) task during training. Two versions of the task we re compared: integrated (e.g., the word RED in the color green) and se parated (e.g., green asterisks above the word RED). Stimuli were congr uent (RED in red), incongruent (GREEN in red), or neutral (XXX in red) . Over 6-day (Experiment 1) and 10-day (Experiment 2) training session s, facilitation due to congruence was small, stable, and equivalent in both task versions. In contrast, interference declined sharply on the integrated task over Days 1-3, then slowed to parallel the gradual de cline on the separated task. Finally, training on the color naming tas k did not affect a word reading task administered after training. Thes e findings imply that (a) Stroop interference initially reflects two p roblems-overcoming integration and managing two conflicting informatio n sources; (b) with practice, the larger integration problem is solved relatively quickly, rendering the integrated and separated tasks quit e comparable thereafter; and (c) facilitation and interference in the Stroop task may be independent. These results challenge extant theorie s of the Stroop effect, which do not predict such effects.