Virtually no evidence for virtually perfect time-sharing

An examination of previous claims for virtually perfect time-sharing in dual-task situations reveals confounding effects that may have obscured dual-task interference. Two experiments are conducted in which these confounding effects are minimized, revealing statistically significant dual-task interference. These results support the hypothesis that human information processing is dominated by a structural central capacity limitation and call into question the hypothesis that dual-task interference can be eliminated by meeting the 5 conditions outlined by D. Meyer and D. Kieras (1999).