| Abstract: | One version of requivalent, calculated from Fisher's exact test p values and recommended for small samples, is considered "a more realistic . . . [and] a more accurate estimate of the population correlation than . . . the sample correlation, rsample" (R. Rosenthal & D. B. Rubin, 2003, p. 494). Small sample properties of rsample and of two effect size estimators (requivalent* and rhybrid) that use requivalent were examined: rsample is preferable to requivalent* (defined as requivalent used without restrictions) in terms of bias and mean squared error (MSE); rhybrid (defined as requivalent only when rsample = 1.0) is generally preferable to requivalent*, and preferable to rsample in terms of MSEs, except when population correlations are very large. Conditions favoring rsample over requivalent* and rhybrid in meta-analyses are noted. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
