Effect size estimates in chemical aversion treatments of alcoholism

Abstract

This study found that aggregate studies on alcohol aversion therapy tended to support a moderate level of treatment impact that may have noteworthy practical import. Emetics appear to generate fairly consistent findings; a paralysis‐inducing chemical may produce variable results.

Effect Size Use in Studies of Learning Disabilities

“…This description led to Friedman's description of effect size measures for most common significance test statistics, including t, z, F, and chi-square (χ 2 ). Thurber (1985) and Good (1983) described some of these same statistics. Friedman (1968) also gave a generalized description of the correlational effect size measures derived from these test statistics: the square root of the test statistic value divided by that same value plus some measure of sample size, typically N or df.…”

“…Studies that fail to find statistically significant results are less likely to be published than studies that include statistically significant results (Schmidt & Hunter, 1995;B. Thompson, 1996;Thurber, 1985;Wampold et al, 1983;Woolley & Dawson, 1983). Any analysis of data without attempting to correct for this publication bias is likely to yield inflated levels of effect size.…”