Re: Clinical vs statistical significance

[truhons]

It is quite possible for a study or a set of studies not to reach
statistical significance because of small n's. Yet when combined in a
meta-analysis there is a significant effect size. I heard Frank Schmidt
talk about this advocating the removal of statistical significance and
replacing it with confidence intervals. I understand the APA is 
encouraging researchers to report more than just significance but effect
sizes or confidence intervals as well.

Stephen Truhon
Department of Social Sciences
Winston-Salem State University
Winston-Salem, NC 27110