Consequences of relying on statistical significance: Some illustrations.

Researchers and journals should abandon statistical significance as a pivotal element in most scientific publications. Confidence intervals around effect sizes are more informative, but should not merely be reported to comply with journal requirements.