|
Topic
|
Power Analysis |
|
Authors
|
George A. Kenna & Joseph S. Rossi |
|
Title
|
Statistical Power in the Journal "Experimental and Clinical Psychopharmacology". |
|
Source
|
Presented at the 73rd EPA Convention in Boston , MA., March 8, 2002. |
|
Abstract
|
The issue of statistical power is largely ignored by many psychology journals including those specializing in lab based research. Objectives: (1) Estimate power in Experimental and Clinical Psychopharmacology" (ECP) and (2) estimate power using the average power by study and congregate average of power given each statistical test. Method: Power was calculated for 619 statistical tests in 48 articles from the journal "Experimental and Clinical Psychopharmacology" (ECP) in the issues from February 1999 through November 2000, inclusive. Results: Power to detect small, medium, large and "extra" large effects was determined for the two year period, and yielded power to detect effect sizes of .12, .36, .59 and .72 respectively. Conclusion: The studies examined in ECP appear to have inadequate power to detect all but the most robust "extra" large, effect size. |